WORLD METEOROLOGICAL ORGANIZATION INTERGOVERNMENTAL OCEANOGRAPIllC COMMISSION (Of UNESCO)
MARINE METEOROLOGY AND RELATED OCEANOGRAPHIC ACTIVITIES
REPORT NO. 26
THE ACCURACY OF SHIP'S METEOROLOGICAL OBSERVATIONS
RESULTS OF THE VSOP-NA
WMO/TD-No. 455
1991
( NOTE
The designations employed and the presentation of material in this publication do not imply,the.expression of any opinion whatsoever on the part oftheSecretarjats ofthe World Meteorological Organization and the Intergovernmental Oceanographic Commission concerning the legal status of any country, territory, city or area~ or ofits authorities, or concerning the delimitation ofits frontiers Of boundaries.,
Editorial n,ote: This publication is an offset reproduction of a typescript submitted by the authors and has been produced without additional revision by the WMO and IOC Secretariats. THE ACCURACY OF SHIP'S METEOROLOGICAL OBSERVATIONS
RESULTS OF THE VSOP-NA
Kent E.C.l) Truscott. B.S.2)
Taylor; p.K.l. and Hopkins, ].S.2
1 JamesRennell Centre for Ocean Circulations ChilworthResearch Park. Southampton) UK
2 Meteorological OfficeJBracknellJBerkshire) UK.
PREFACE
~t...eorol(>gical c..>ooervatiC)i16 ma.de onlx>ct.rd. merchant ve(~se18 of the WM(J ·VOlurlt..-:3.ry observtrlg SliiJ.)6 (VOS) e'JCheme, when transmitt-..ed to s}10re in real-tilfJe., are :t1. t~ll.ootml.tial (~()I'm;JOn.ent elf ttle Gloool ObBerving System of the World. Weattter Wat~h ~to marine~teorol~ical and. are E:Ssent;ial -the provision of 6erv-ices.J aB Wf"~l.l as to meteorological analyses and forecasts generally. TtleBe observations are c.\lBOn~~rcl€d in ahips' lJ)9u::orological l<:gbooks, for lat.,er excharlge, arctliv;:il arld. l)l~'O(~~tJirJg ··ttll'"()1.Jg}l the WHO M<.u-ine ClimatologiC'a..l Stmmaries Sch~, an
tile climc::rt()logy ()£ t~tle u.c.:,r"ine atJD(X:il)t,'l<:~r(?: rnad (X-~(~;...lJl sllr'f;;l(-x::,~1I ,t~ul(i ·.f()r (~()mpItiI:Jg a vari.t:~ty of c.\ir·~-6fXl flux.cc:; ~ At th(~ ~time, however, it llaS loragbeerl rect)f:.trlized. that tJlef>e oooerv(ltioIlB are 6l1bject to errc>rB, both 8YSt;f~t~ic and random~ Many ()f these (~rro'rs clr(:; the l.'e6ul't of inadequate o·t:.. inappropriate irtf)t~rument siting Olllx);:J.:rd ship, ,or t,hro.Jgh tIle UBe of instrumerrtatior1. or ()bc'~~rving t.,,(~hlli(lues wJile!:.. ~ less ·t!k1l1 opt;imt1.1.
Th.e V(JS SJ:«;ial Ot'l3f~DrirJg PtX)je<;t~ Nortll Atlmltic (VSfjP-NA) was therefoI-e irlitiat-.-e:l, jc>in"tly by 'tile WHO ('iOJIID'jission 'for Marine .~teor'()logy (aiM) arld ttie ('1a.ngf:~~ CoIrmittee ·0[1 Cli.l1k:lt.e arJd the CCean (COX» ()f IOCjsaJR.J on behalf of'th.e WCRP, to try tA) establif:;ll the E~ffec-tB 011 the qlJalit~y of vas (]a:U1. of diffen~nt, sli.iJ:) inrrt;'rtJ.J.nerrta.1;i()ll 811(1 clbt3e:rv'il'ig I)rac.:tiet::i3 ..
Si.xllati.orlf..t.l otJf)erviIlg fl_eet.:.t1 pcrrt-icipatal .- tllo...ere c,f (~«lad;.l.l Fran.c"e, f'~·.rmc11.1Y., Net~llerlands.l fJrli·tA:Xl Kingijom ;:.rlfl USA ._,. arl(j l.l1timcft-ely 45 ships 8l1I~li(:'rl cL:l:t;a '~for t~h(~ pr-ojt::~1; ~ New loglxx>ks we~ desigrled. -to enat)le t~he 8Ct}IJ.i.sitioll of tllll~pl(~lnent..ary in,f()rmat~ic111 "fA) (le-fine 'the det..ailetj instrl.1JDentation and practi..c"(":S ill llS(~ by f.Xtch alii!), t30 -tb..at the E}ff~ta of -trl~ differing me·tclods of data g.rxtller~ing c;()1.ilcl lx~ q1.U"3rrt.i~fi.e Mf;}tt:~j·r(,)logi(;a.l ()ffi(;e ill Brack.rlell If)r f1n-.:hiva.l GlIld f!rJalysis J j()irltly by tt-le ~t~).r."()lOf?:i(~~.l Off1.c:e [jI1C! tht::~ ~JaJJ.i(~ Rerlnell C€~rlti're in So'u.tllaiIiptnn. Eventually a t~'rtal of lOC>re than 33,000 ooc*~rvatior16 we~ (-;ollected dllri:r.g -the project ()lx:.~(~rvat..i()rl Pf.~ricxl frc..>m MEty 1988 to Sept..elflber 1990 andt.he-..se, t,c,ge"tllf=..:r wi1-11 thE~ irl.fol1DcftioIl on l.llS·trtJlfJf;Tlt si-ting al1d expt;)t')ttre and thf.': mt~teotx)l()g·i.c:al an.alysis 'fields f:rrJD'lthe Ilume,ric:.r3.1 JOO(Jel of thf~ Uni.t."ed Kingdom Meu.;ol'X)l(Jgical Offi(~.~ provi Tllis P;Jrt:.icll.1ar~ (j TI'lf~~ is :O() d.OlJ.bt 'th-'3.t~ tIle re6l11t..~ of 'this prr>jE:Ct are of (X)r}Bide~ratJle ill)"POrtall('~ -to (;lmlt.(~ allEtlysis and DYJClell.il:1g~ ill pm.-tic~lar III tllf.~i.r impllL~}tions for t~tle (~()n~puta..tion ()f a.,i·r---·f.*Xl flt1Xe,s of h(Xit~ UK.>merrtum at1d water V~pt.)l.ll·'.. TIley a.r~ als() li..kf~l·y· fA> haV(~ a 6ignifi(~t and. be'i1.efi(~l irnp;-l.ct evc~rlttu:illy t,")rl tJit.:1 c~perati(JIl of th~~ wb.oIe of tIle VOS, with C()DBeq1.1erJ.t bene'fits not jlJ8t~ f():r re6f..}BlY~tl bu.t also for" oj?erati()rial mt~t,eorol()gy~ vi trhe considerable f.f~pprecia·tj~()rlofthe SprJll130rir.ll~ ()llIarrizcltio:rlB :for· t~:tl(:} ~~xt~Jrlded ··th(~allt;11()n1 ·thf::;S£~ E.Kt~rl.t~ project· is ·tl() ()-.r. -t'wo repcjrLs.t Ms arid. Dr .. ·P.~.Tayl()r ()f tJle J;.m.l(~ Rennel,l r~llt:r~IDld Mr.B.. BnlB(~ol;t~ arld.Mr." ~J "HopkitlB of,:1~l1e.Unit~jKirJg(lom Me~t:.E:~'rYJlog:i(.k~10ffic~foI·thel:r sua:rt...arltial.·lirld 111.f:~1 <}lm-l,i1:iY ;:lllalYBi~J work.. Thmilu)aT(~ a.lso clue fA> ·t~he ~(~ttle'lanft~,·;Harol~lrg,for \JncleI-t,.t1king 1:~lle m;:u()r t~iJ,£:)k ()f ·(11g:i:tizinglJledaUt; tXJ t;Jlf::=fnE~l:en3 ()f t~ll(':~ I)n)j~~~t M;Iftaf.!l~fllf:.?;rlt C:()JJlnit~t~)(~ 'f()t~ t;hf~1.'r (:::xc"(::11('~Tlt 8l~pervi.si(.)n ()f t~lle Pl:\)je<;l;; to, 'tlleP(")rt Metk·:~()t~ol(.>g;ir~al·0f·fi(~r6()f t.lle (x)tHltries (~)Il(~rn£~j fo~r' n:X;'rI.li-titJg; Gl.:r.irl aervlt.::irlg t~l}(~~ .. Ill.ujec:t, r;}lif)f.);.arrl, 1<:u3t~ ,.l:llt· by', 'l'l() mt~lB It:xmt, lJ()tll~~ ():ffl(~r~li~; a.Ill1 ·crew of "ttl{~ alliJ)S' :tJ"f':mlBe],V{~,-for" "t:J)(~ir(:()"---()P:~I"ilti()Il.,;lr.(') fitlPfl(),t~'t ,"f()l"lJl(:; P·lY..>j~~t'J wi··tlI.Ollt Wllich ,I)()iihil}J5'w(Xll. ."). '7t#4/fr1,(~/7.lft~,:,4f(LVt.r6 .. (&. .Kul.lerlDA€, ) I &~:,~'[¥~~t.(:.~ry· lex:: " l Contents Page l I EXECUTIVE SUMMARY 1 t 1. INTRODUCTION 3 1.1 VSOP-NA project 3 1.2 The need for augmented observations 3 1.3 Method of analysis 3 2. THE VSOP·NA DATA SET 4 2.1 Data Acquisition 4 2.2 Instrument codes 5 2.3 Ship Routes 8 2.4 Editing of the VSOP-NA data 8 2.5 DataArchiving 10 l 3.1 Choice ofthe Model 12 3.2 Calculation of the Analysis fields 12 3.2.1 Introduction 12 I 3.2.2 Sea Surface Temperature 13 3.2.3 Mean Sea Level Pressure. Wind Speed and Wind Direction 14 3.2.4 Temperature and Humidity 14 3.3 Model Treatment ofthe Boundary Layer 15 4. STATISTICAL SUMMARY OF THE ANALYSIS RESULTS 15 4.1 Air Temperatures 15 4.2 Dew Point Temperatures 15 4.3 Sea Surface Temperatures 17 4.4 Wind Speeds 17 4.5 WindDirections 17 4.6 Pressures 17 5!1 INTRODUCTION TO THE DETAILED ANALYSIS 17 6. SEA SURFACE TEMPERATURE (SST) 18 6.1 .Range of SST Values Encountered 18 6.2 M,an SST Differences - Ship by Ship 19 6.3 Variations due to Recruiting Country 19 6.4 Variations due to Instrument Type 19 6.5 Effect of Ship Size 20 6.6 Summary· SST 20 7. AIR TEMPERATURE AND HUMIDITY 21 7.1 Introduction 21 7.2 Range ofValues Encountered 21 r r r t l viii 7.3 Geographical and Seasonal variations 21 7.4 Variations by Ship and Recruiting Country 22 7.4. I Air Temperature 22 7.4.2 Dew Point Temperature 23 7.5 Variations due to Instrument Type 23 7.6 Effect ofInstrument Exposure 23 7.7 Summary. Air Temperature and Humidity 24 8. WIND OBSERVATIONS 25 8.1 Range ofWind Values Encountered 25 8.2 Geographical Variations 25 8.3 Wind Speed and Direction • Ship by Ship 25 8.3.1 Wind speed 25 8.3.2 Wind Direction 26 8.4 Variation with Wind Velocity 26 8.5 Variation due to Method of Estimation 26 8.6 Anemometer Wind Data 27 8.6.1 Variation with Recruiting Country 27 8.6.2 Effect of Anemometer Height 27 8.6.3 Variation with Instrument Exposure 27 8.6.4 Errors in estimating the True Wind 27 8.7 Visual Wind Data 28 8.7.1 Variation with Recruiting Country 28 8.7.2 Comparison of Day-time and Night-time values. 28 8.7.3 Beaufort Scale Effects 28 8.8 Summary. Wind 29 9. PRESSURE 29 9.1 Range ofPressure Values Encountered 29 9.2 Geographical Variations 29 9.3 Mean Pressure Differences· by Ship, Country, and Instrument type 29 9.4 Variation with Mean Pressure 30 9.5 Barometer Height Corrections 30 9.6 Summary· Pressure 31 10. SUMMARY 31 10.1 Model Characteristics 31 10.2 Instrument Characteristics 32 10.3 Augmented documentation 33 11. RECOMMENDATIONS 34 A : Recommendations To Members OperatingVOS 34 B: Recommendations to WMO 35 C : Recommendations for Climate Research 36 12. ACKNOWEDGEMENTS 37 13. REFERENCES 37 -1- l I t I EXECUTIVE SUMMARY The effect on the quality of Voluntary Observing Ship (VOS) data due to different t instrumentation and observing practices has been evaluated during the Voluntary Observing Ships l, Special Observing Project for the North Atlantic (VSOP-NA). Forty-five VOS ships operating in the North Atlantic were recruited into the VSOP-NA by the participating VOS operators (Canada, France. Germany (FRG). Netherlands. United Kingdom and United States). For these ships. f additional information was obtained on the instrumentation carried (Kent and· Taylor, 1991, Marine Meteorology and Related Oceanographic Activities. Report No. 25. WMO Geneva) and on the observing practises used (reported on logsheets with each observation). The output from the analysis phase of the U.K. Meteorological Office Fine Mesh atmospheric forecast model was used asa standard to compare one ship observation with another. Important results include detection of the effects of model characteristics. of biases due to the. different types I. of instrumentation used. and an assessment of the value of the additional data groups reported. l a) Model Characteristics. Tiie VSOP-:t~A results showed the value oi L"1e ship l observations for quality controlling model analyses. Compared to the ships, model I values of air temperature ',a/ere fOUIld to be higher in colder·regions a~"1d 'vice versa. Particularly in winter there was a geographical trend with the model temperatures increasing to the northwest compared to the ships. For humidity the model values were drier under low humidities but moister for the most often observed humidity values. Model derived surface wind speeds were biased low by an estimated 2 knots l compared to the ships. Airpressure comparisons showed a north-south trend in the I model values of about 0.3 mb. r b) Instrument Characteristics. The use of different measuring techniques v·/as found to produce significant differences in the data characteristics. Engine intake r derived sea surface temperature values were warm by about 0.3°C compared to hull ( sensor values, and, except in. low· heat flux, stable conditions, compared to buckets. Psychrometer and screen derived air temperatures generally· showed similar errors due to solar radiation (of order. one or two °C, the effect being modulated by the wind speed); but some screens had particularly poor exposure and correspondingly larger errors.Psychrometer derived dew point temperatures were lower (and therefore more likely correct) compared to screen values by r between 0.5°C and 1.5°C. Even after allowing for the height of measurement. the fIXed anemometer readings were greater than Beaufort estimates; furthermore there l were many errors in correcting measured winds for the ship velocity. Pressure values from digital Precision Aneroid Barometers were less scattered than those from analogue instruments. I c) Augmented Documentation. The major cause of degradation of the data, {. detected by the additional data groups provided with. each observation,was the t, correction ofmeasured relative winds to true winds. Where more than one method t of sea surface temperature measurement is used the measurement method must be known· to remove possible biases; this is not· presently the case for observations ( transmitted by radio. r r r -2- On the basis of these results a series of recommendations were'formulated by the VSOP~ NAManagement Committee. These included1: a), Recommelt'dations to' members" operatingVOS .. concemedthe nee-d for' improved observing practices and equipment. .the need to publicise fue VSOP-NA results,ancd the desirability for·increased .. real-time data monitoring. .Errors· in true wind and dew point calculations ,should be decreased by the provision·of dedicated computer programs 'or calculators. ti) aecommendations to ,WMO concerned the reporting of sea surface temperature ,measurement. method .and of the/ships draft or "height of eye";. possibly.requiring the 'implementation of coding changes. A need was identified for improvement 'ofthe List of Selected. 'Ships 0NM047) and also fora reference booklet on observing practices. The irnp0rtance of the Port Meteorological.System to data quality was emphasised. TheVSOP NAresults should be considered in re-assessment of the Beaufort Scale. andconsideration given to the reporting of relative wind in ships' logbooks. e) aecommendations for Climate,Research concerned the use of vas dat.a to verify model flux determinations. that sea surface· temperature data should be processed separately according" to observation method,and that support should be given for measures to improve the Port M,eteorological Officer system. For the full text of these recou.unendations see Section llaftile report. - 3- 1. INTRODUCTION 1.1 VSOP·NA project The Voluntary Observing Ships Special Observing Project for the North Atlantic (VSOP-NA) (WMO. 1987. 1988) was set up to establish the effect on Voluntary Observing Ship (VOS) data of different instrumentation and observing practices. Six national observing fleets participated - those of Canada. France. Germany (FRG). Netherlands. United Kingdom and United States - and the aim was to recruit about 50 merchant vessels with good observing records to provide extra information to supplement their conventional observations. In the event. 45 vessels supplied data. A catalogue of the VSOP-NA ships (Kent and Taylor. 1991) has been prepared describing in detail the ship characteristics. routes and meteorological instruments fitted. This document provides a summary of the data acquisition. data processing and ·analysis phases of the project. and describes the results of the analysis1. \Vhile maiiY of these errors are likely to· be small and perhaps i.l,significant for operational forecasting. even small biases can have significant effects in air-sea flux estimations. With the increased requirement for longer term ·forecasting and for ·climate research and prediction; identification of these errors. through projects such as-VSOP-NA is now urgent. 1.2 The need for augmented·observatioDs It has long been recognised that observations made on board merchant vessels are subject f to errors. both systematic and random. There are clear difficulties in siting instruments so that ( they provide measurements which are representative ofthe I free I air undisturbed bythe presence of the ship itself. Some national services consider that estimates of wind speed. made using the ~ Beaufort Scale. are preferable to measured values which are subject to unquantifiedcalibration. exposure and reading errors. Also. the design and siting.of· thermometer screens· varies' widely r between national, services. and some recommend the use ·ofpsychrometers to increase the r likelihood of ·adequate ventilation. Data submitted by theVOS both in· real time and later·via logsheets ,include only very f limited infonnation to allow discrimination between different' observing practices. A major aim of r the VSOP-NA'was to acquire supplementary information to define the detailed instrumentation and I practices in use by each ship. so that the effects of these differing methods of data-gathering could f be quantified.' \ ( 1.3 Method of analysis The acquisition artdassemblyoftheVSOP-NA dataset is described in Section 2. To provide a reference against which observations',could·be::compared. archived analysis,·fields provided by a f 1 Preliminary reports were prepared during the analysis phase by Starkey (1989) and Truscott f (1990). In addition. a total· of.seven newsletters were circulated from theUK archiving centre to r keep the participants·infonned of progress with data collection and early analysis. r f) ( -4- Numerical Weather Prediction model were accessed, and each VSOP observation matched against its corresponding model analysis value for the same time and location. Differences (observation minus model) were summarised by ship, by observing practice, and by prevailing conditions, to expose these effects in quantitative terms. The model analyses, while, being subject to various errors due to simplifications and assumptions in the model physics, 'can be regarded as providing a useful reference datum. Thus, we are not considering the absolute accuracy of the model's performance (although'some conclusions can be drawn, see section 10). Rather, using the model, we are seeking to compare ship observations from widely differing, areas and times in order to expose relative differences arising from the VSOP ships' differing instrumentation and observing practices. A problem in this approach is that certain ofthe VOS observations have ,been used in the model initialisation scheme; this will. be considered in Section 3, and in discussion of the, ,affected variables. A statistical summary of the analysis results is presented in, Section 4, followed by a more detailed discussion in Sections 5 - 9. The results show that statistically significant differences exist both between and within national observing fleets; that some observing methods appear more consistent than others, and that certain biases can be linked to inadequate exposure of equipment. It is hoped that these results will. assist in establishing the inherent accuracies of the VOS'data, for application in such areas .as 'operational"weather forecasting, climate change studies and satellite data verification. In the light of these results (summarized in Section I0), some recommendations are formulated (Section 11) regarding observational practices on board vessels of the Voluntary Observing Fleets. 2. THE VSOP-NA DATA SET 2.1 Data Acquisition The identification and recruitment of suitable ships was the first difficult and 'time consuming phase' of the project. An initial selection was made on the basis of the reporting perfonnance of ships regularly plying the North Atlantic. Actual recruitment was undertaken by Port Meteorological Officers (PMOs), using a presentation package prepared centrally. Since, the additional information· on instrumentation and observing practices was, crucial for the success of the project, every effort was made to obtain detailed and complete descriptions of instrument types and exposures. The acquisition of this information continued throughout the data collection phase, and ultimately a catalogue was assembled (Kent and Taylor, 1991) containing the fonowing information for most vessels: i. Methods of observation of wind, temperature, humidity, sea surface temperature and pressure. n. Height of all instruments above a nominated reference level. in. Ships' plans clearly showing the instrument positions. iv. Photographs of the instnnnents showing their exposures. v. Limited model/manufacturer details for equipment in use. I - 5- f f With each observation. the following supplementary information was requested. to be r entered in columns added to the standard logsheet format: i. Ship's speed. l ii. Ship's heading. iii. Height of deck cargo above sealevel. l iv. Height of nominated reference level above sea level. v. Method of sea surface temperature measurement. vi. Location of air temperature measurement. l vii. Relative wind speed. viii. Relative wind direction. The first VSOP observations were made (by Dutch recruited vessels) in May 1988. but most Gennan and UK recruited ships started VSOP-NA observations in September of that year. and USA and French· ships in the· early months of 1989. Table la shows the number of observations ( submitted by eachfor each month of the project. I, Observation sheets were collected by the responsible .national meteorological service and forwarded to the data keying centre at Deutscher WetterDienst. Hamburg. The data format used I. I was an extended version of the IntemationalMarine Magnetic Tape (ThAMT) format. with the supplementary information added at the end of each record. Magnetic tapes. were then diSpatched to the archiving centre at Bracknell every two to three months. This procedure worked very satisfactorily. although there was often a time lag of several months between data being.recorded and receivectat the archiving.centre. The project timetable required that the fmal analysis be complete by summer 1991. and so the archive was 'frozen' in late. 1990 after data for September 1990 were receiveci. The total number of observations archived in the observation dataset was 33.6k. To allow comparisons to be made with data· originating from a smaller vessel manned by professional observers. data from OWSCumulus. when on station LIMA (57°N. 20°W). were collected and added to the· archive. About 3.6k observations were obtained from this source. The original aim of acquiring at least a full annualcycle.of observations was achieved by 28 of the 45 vessels. with one vess-elsubmitting 25 months of observations and another 24 months. In contrast. 8· vessels submitted data over a period of 6 months or less. Sadly. one vessel. the sole Canadian participant f Irving Forest. was lost at sea in January 1990; fortunately all hands were safely rescued by a r passing ship. r Although the data received fell short· ofthe original plan of at least 12 months of data from SO ships. recruitment and retention of vessels posed greater problems than had been expected; r! also. full vessel descriptions and instrumentation details proved difficult to ·acquire in some cases. r However. this· report will show that a satisfactory database has been collected to allow meaningful r conclusions to be drawn from the analysis. 2.2 Instrument codes f To represent the extra information on·.instrumentation. special codes had to be defmed r (Tables 2a -g).. Since participating countries· use differing codes to denote the method of sea r surface temperature measurement. ·.the codes submitted by the ships were converted into a standard code (used for all countries) shown in Table 2c. Codes representing the location of air I ) I I - 6- Table la Calendar showing the number of observations submitted each month by participating ships. SHIP 88 89 90 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 1112 C60S 0 0 0 0 0 0 0 0 0 0 o 86 27 32 68 49 39 49 19 32 45 39 23 74 28 0 o· ··0 0 0 0 0 0 0 0 0 610 VSBV3 0 0 0 0 0 0 0 o 62 68 70 78 99 75 5582 64 65 94 69 70 o 75 59 77 42 56 64 60 69 32 0 0 0 0 0 1485 GJ"R 0 0 0 0 0 0 0 o 37 72 96 8761 83 93 82 70 55 46 47 63 63 79 68 75 72121101109 99 32 0 0 0 0 0 1711 VRAZ 0 0 0 0 0 0 0 7 48 63 70 46 60 37 38 o 56 55 69 76 41 82 76 57 66 73 55 73 6645 0 0 0 0 0 0 1259 GXES 0 0 0 0 0 0 0 0 o 69 68 72 66 44 70 64 78 72 41 74 54 69 91 52 93 54 59 28 33 8 31 0 0 0 0 0 1290 GJ"S 0 0 0 0 0 0 0 0 7.90 72 60 52 3967 58 56 59 60 50 76 77 73 59 71 61 59 15 74 69 .0 0 0 0 0 0 1304 GBVV 0 0 0 0 0 0 0 o 52 38 67 91 54 6868 51 18 30 43 42 68 84 53 57 76 67 80 58 68 57 n. 16 0 0 0 0 1378 GZHH 0 0 0 0 0 0 0 o 102943 47 21.52 5043 44 56 37 47 14 29 43 32 52.33. 44 53.47 37. 0 0 0 0 0 0 863 GBSA 0 0 0 0 0 0 0 o 29 64 44 39 43 45 31 35 51 22 8 0 28 8 52 29 29 26 39 69 31 27 0 0 0 0 0 0 749 GBVV 0 0 0 0 0 0 0 o 39 66 50 40 82 93 64 60 64 60 97 72 58 56 49 33 27 454 73 83 71 71 0 0 0 0 0 1366 DOLH 0 0 0 0 0 0 0 o 30 1365 57 53 36 57 6245 48 56 19 28 52 42 25 12 11 10 0 ..· 0 0 0 0 0 0 0 0 787 ODUC 0 0 0 0 0 0 0 0 0 o 34 66 32 46 19 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 197 DHNE 0 0 0 0 0 0 0 o 49 46 41 56 55 46 41 45 40 42 14 44 4648 53 61 8740 54 9349 0 0 0 0 0 0 0 1056 ONBR 0 0 0 0 0 0 0 0 o 58 81 69 94 86 51 20 4 0 0 0 0 0 o 0 o 0 0 0 0 0 0 0 0 0 0 0 469 OHRG 0 0 0 0 0 0 0 0 0 2 68 53 48 70 66 42 52 34 3 32 56 56 46 69 66 32 38 17 0 0 0 0 0 0 0 0 850 OI"C 0 0 0 0 0 0 0 0 o 30 53 46 30 51 41 28 10 49 60 43 35 43 55 32 070 43 .. 63 31 0 0 0 o. 0 0 0 813 DNJR 0 0 0 0 0 0 0 o 24 32 39 36 3721 42 11 37 37 49 1 53 20 38 31 35 11 29 33 4 0 0 0 0 0 0 0 626 PCEl 0 0 0 0 0 0 0 0 0 o 38 62, 16 63 24 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 263 PGOS 0 0 0 0 0 0 0 039 36 40 35 ·46 26 68 42 5966 66 42 11 51 43 6138 5413 0 0 0 0 0 0 0 0 0 902 PGDG 0 0 0 0 0 0 0 0 o 41 61 11 5026 63 68 41 54 62 60 72 74 3464 50 71 32 49 17 0 0 0 0 0 0 0 1006 PELT 0 0 0 o 31 50 36 60 29 0 0 0 o 0.0 0 0 0 o 0'0 0 0 0 0 0 00 0 0 0 0 Q 0 0 0 206 PElU 0 0 0 o 5044 40 51 26 00 0 o 0 o 0 0 0 o· 00 0 o 0 0 00 0 0 0 0 0 0 0 0 0 217 PGEG 0 0 0 2 31 35 2161 32 31 2928651841 29 28 40 20 59 22 3824 4525 41 16 8 0 0 0 0 0 0 0 0 789 PGOY 0 0 0 0 0 0 0 0 0 0 o 18 12 60 19 43 28 3031 26 61 29 3148 31 4130 30 31 0 0 0 0 0 0 0 599 WPYH '0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 32 40 51 43 6048 43 6148 38 53 0 0 0 0 0 0 0 0 0 0 517 WPVl 0 0 0 0 0 0 0 0 0 o 0 0 0 o 39 44 57 45 40 56 61 58 31 56 61 39 62 11 0 0 0 0 0 0 0 0 672 KRGJ 0 0 0 0 0 0 0 0 0 0 0 0 0 o 47 o 55 4186 1 35 23 41 53 59 21 33 0 O' 0 0 0 0 0 0 0 501 UPFZ 0 0 0 0 0 0 0 0 0 060 0 0 o 53 64 43 o 56 55 50 31 57 21 0 0 0 0 0 0 0 0 0 0 0 0 496 UPHZ 0 0 0 0 0 0 0 0 0 0 0 0 o 72 7667 67 72 66 65100 70 64 72 94 41 0 0 0 0 0 0 0 0 0 0 932 V"lG 0 0 0 0 0 0 0 0 0 0 0 o 10 41 5843 35 25 33 22 02320 0 o 0 0 0 0 0 0 0 0 0 p 0 310 OXTZ2 0 0 0 0 0 0 0 0 0 0 0 0 0 o 51 69 7113 24 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 288 OYGKZ 0 0 0 0 0 0 0 0 0 00 0 0 00 0 o 14 18 16 15 16 13 20 0 0 0 0 0 0 0 0 0 0 0 0 112 VSDG 0 0 0 0 0 0 0 0 0 o 17 0 0 0 0 0 0 0 0 o 0 75 39 40 o 22 0 0 0 0 0 0 0 0 0 0 19.3 KRJL 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 81 66 595143 64 38 54 69 8 34 10 4939 0 0 0 0 0 0 0 665 KRU 0 0 0 0 0 0 0 0 0 0 0 0 0 7 48 57 56 3641 1671 56 39 49 o 0 o 0 0 0 0 0 0 0 0 0 542 KRJP 0 0 0 0 0 0 0 0 0 0 0 0 o 35 65 52 66 12 60 6214 0 o 0 0 0 0 0 0 0 0 0 0 0 0 0 426 KRPS 0 0 0 0 0 0 0 0 0 0 0 0 0 062 5434 47 49 49 48 60 58 56 49 20. 0 0 0 0 0 0 0 0 0 0 586 FHFD 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0111182153180169184192111184193158178180177183178182162 0 0 0 3123 FNEf 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 26 34 24 39 53 49 25 35 24 16745 5060 53 41 39 67 0 0 0 . 732 FHGS 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 14 017 12 0 0 0 o 11 12 4 0 0 0 0 00 0 0 0 0 70 FHGH 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 31 37 41 23 0 0 00 0 0 0 0 o 26 28 24 15 0 0 0 0 231 OIOA 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 95 69 13 17 8277 54 29 0 0 021 0 6 o 70 94 0 0 0 0 631 FNCZ 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 32 6542 13 o 27 54 70 50 56 53 16 1 65 68 67 0 0 0 0 0 679 FNOY 0 0 0 0 0 0 0 0 0 0 0 0 0 0 o 0 0119205153 66 72 0 041 0 0 0 0 0 0 0 0 0 0 0 656 VSBG8 0 0 0 0 0 0 0 0 0 0 0 0 o 43 50 45 65 36 46 29 58 70 46 0 0 0 0 0 0 0 0 0 0 0 0 0 4sa OVERALL TOTAL -33645 Table Ib Calendar showing the number of observations submitted.each month by participating ships which were matched with model values. SHIP 88 89 90 1 2 3 4 5 6 1 8 910 11 12 1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9 10 11 12 C60S 00000000000 11 27 31 52 45 35 36 3 2 43 37 22 13 28 0 00 0 0 0 0 0 0 0 0 511 VSBV3 0 0 0 0 0 0 0 0 61 62 6818.99 15 51 81 58 57 90 12 70 0 12 51 12 42 56 63 60 66 31 0 0 0 0 0 1375 GJHR 0 00 0 0 0 0 0 30 50 18 13 47 65 75 65 48 39 33 8 45 le2 49 55 55 52 49 68 69 67 22 0 0 0 0 0 1184 VRAZ 0 0 0 0 0 0 0 1 1438 35 18 19 16 18 025 35 28 8 19 35 41 3424 32 16 36 37 17 0 0 0 0 0 0 552 GXES 0 0 0 0 0 0 0 0 069 66 12 66 44 68 63 75 69 41 1 54 66 87 48 80 54 35 28 21 0 22 0 0 0 0 0 1141 GJHS 0 0 0 0 0 0 0 0 7 62 45 36 31 28 49 40 41 39.42 3 50 60 41 43 53 48 22 2 59 55 0 0 0 0 0 0 868 GBVV 0 0 0 0 0 0 0 03629 48 72 3854 5143 13-25 21 13 446242 45 57 44 56 43 5143 55 16 0 0 0 0 1001 GZHH 0 0 0 0 0 0 0 0 1029 42 47 21 5148 le3 42 52 37 36 10 21 43 31 50 32 24 50 47 37 0 0 0 0 0 0 809 GBSA 0 0 0 0 0 0 0 0 18 44 123121 32 25 22 31 11· 8019 0 36 23 16 20 10 49 ·1 27 0 0 0 0 0 0 456 GBW 0 0 0 0 0 0 0 0 28 49 36 30 49 61 53 41 31 46 16 29 45 3631 25 13 4 3157 56 55 50 0 0 0 0 0 944 OOLN 0 0 0 0 0 0 0 0 30 73 62 5153 35 55 6143 4455 4 28 50 42 2412 17 6 0 0 0 0 0 0 0 0 0751 POUC 0 0 0 0 0 0 0 0 0 0 23 50 29 21 19 0 00 0 0 0 0 0 0 0 0 0 0 00 0 0 0 0 0 0148 OHNE 0 0 0 0 0 0 0 048 45 41 56 55 45 38 44 37 42 14 8 45 41 49 59 82 39 54 91 42 0 0 0 0 0 0 0 981 OHBR 0 0 0 0 0 0 0 O. 0 56 n 68 94 84 56 20 4 0 0·0 o. 0 .0 0 0 0.· 0 0 0 0 0 0 0 0 0 0 459 OHRG 0 0 0 0 0 0 0 0 0 2 61 47 48 67 61 31 48 32 3 3254 50·45 54 51 32 31 11 0 0 0 0 0 0 0 0 712 DIHC 0 0 0 0 0 0 0 0 0 28 50 46 29 50 39 28 941 54 2 27 36 55 25 0 69 35 62 30 0 0 0 0 0 0 0 721 OHJR 0 0 0 0 0 0 0 024 30 36 36 37 20 42 11 33 3446 1 5311 36 31 35 16 28 31 2 0 0 0 0 0 0 0 599 peEL 0 0 0 0 0 0 0 0 0 036 62 76 61 24 0 0 0 0 0 0 0 0 0000000000000 259 PGOS 0 0 0 0 0 0 0 0 2215 13 15 17 17 31 23 40 31 37 16 30 0 33 33 22 20 0 0 0 0 0 0 0 0 0 0 421 PGOG 0 0 0 0 0 0 0 0·0 24 39 022 7 38 37 17 21 2828 40 17 17 24 33 3812 17 13 0 0 0 0 0 0 0 412 PELT 0 0 0 0 19 24 35 46 29 000000000000000000000000000 153 PELU 0 0 0 0 39 27 38 4526 0 0 0 O· 00000000000000000000000 115 PGEG 0 0 0 217 10 8 2823 17 18 21 38 8 29 21 20 30 8 15 15 11·16 24 15 29 7 8 0 0 0 0 0 0 0 0 438 PGO'" 0 0 a 00 0 0 0 0 0 0 12 0 23 1 22 9 12 12 0 21 0 11 22 0 18 11 10 12 0 0 0 0 0 0 0 196 IJPYH 000000000000000 29 36 49 34 0 43 38 61 47 37 53 0 0 0 0 0 0 0 0 0 0 427 WPVF 00000000000000 39 41 51 41 39 37 65 56 37 54 58 37 35 100 0 0 0 0 0 0 0 600 KRGJ 0 0 0 0 0 0 0 0 00 0 0 00 43 0 S2 32 75 ·0 20 .. S42· 1656 80000000000 349 WPF2 00000000000000 43 59 40 0 52 8 49 33 54 21 000000000000 359 YPHZ 0000000000000 63 56 55 45 53 50 37 64 54 49 60 77 46 0 0 0 0 0 0 0 0 0 0709 lJHlG 0000000000009 41 57 43 33 21 33 0 0 19 20 0000000000000 276 OXT22 00000000000000010000000000000000 00 0 0 1 OYGK2 000000000000000000000000000000000000 0 YSDG 000000000000000000000 59 28 2S 0 110 0 0 0 0 0 0 0 0 0 123 KRJl 000000000000000 65 59 41 4210 55 35 41 68- 3 30 4 37 31 0 0 0 0 0 0 0 539 KRLZ 00000000000005 45 51 43 33 39 10 S6 46 38 48 000000000000 414 KRJP 0000000000000 30 57 41 29 53 52 27 11 000000000000000 300 KRPB 00000000000000 61 52 31 40 48 13 42 56 58 55 46 20 0 0 0 0 0 0 0 0 0 0 522 FNFO 000000000000000 21 44 36 3424 35 52 38 54 48 49 52 43 46 54 51 47 38 0 0 0 766 fNEF 000000000000000 2S 28 24 38 22 47 21 34 24 1 62 40 45 56 49 39 38 60 0 0 0 653 FNGS 0000000000000000015000060 4 0 0 0 0 0 0 0 0 0 0 16 FNGH 000000000000000 15 1 17 11 0 0 0 0 0 0 0 0 0 5 11 5 10 0 0 0 0 81 OIOA 0 0 0 0 0 0 0 0 00 0 0 0 0 448 7 6 5 4 30 18 20 0 0 0 5 0 0 034 47 0 0 0 0 228 FHCZ 0000000000000003 11 9 2 0 6 9 13 11 13 7 J 1 11 18 16 0 0 0 0 0 133 FNOY 00000000000000000 91 5 71 58 40 00000000000000 271 VSBG8 Q 00000000000 041 45 45 36 35 46 24 56 66 45 0000000000000 439 OVERALL TOTAL -22592 l l - 7- I l Table 2: VSOP CODES f a Country Codes b Method ofWind Speed measurement I 00 Netherlands 0 Estimated ( 02 United States of America 1 Fixed anemometer 03 United Kingdom 2 Handheld anemometer 04 France 3 Estimated. but with fixed 13 Canada anemometer fitted l 21 Gennany (FRG) Method ofSea Surface Method ofAir Temperature C Temperature measurement d measurement 0 Screen 1 Bucket 1 Psychrometer ( 2 Engine Room Intake 3 Hull sensor I Hull sensor (engine 4 Method ofHumidity measurement f room) e 1" 5 Hull sensor (Met 0 approved) 6 Radiometer 0 Wet bulb in·screen 7 Trailing Thennistor 1 Wet bulb in psychrometer 8 XBT f Exposure Codes (higher figures represent better exposures) 0 fti...rflow fully blocked adjacent to sensor (withi..Tl lm) 1 Airflow fully blocked at medium distance (1 to 4m) 2 Airflow fully blocked at larger distance (4 to lOm) f 3 Airflow partially blocked adjacent to sensor (within lm) 4 Airflow partially blocked at medium distance (1 to 4 m) [ 5 Airflow partially blocked at larger distance (4 to·10 m) 6 Airflow clear with long fetch (> 30m) -over ship ~ 7 Airflow clear with fetch 10 to 30m over ship 8 Airflow clear with fetch 1 to lOm over ship l 9 Airflow clear with short fetch «lm) over ship 9 Location·of Air Temperature measurement f Code Can Fra FRG Neth UK USA KEY: 1 MS P P S IvO MO P- Port I 2 lv1P SS P 1vID MD S- Starboard 3 11F - - - MA MA M- Monkey Island r 4 MA --- ID :so B- Bridge deck 5 BS - - - BK BK F-Forward 6 BP - - - BW BW A- Aft f 7 BF - - - - - O-Outboard 8 BA - - - - - D- Amidships K-Bulkhead r W- Aft of wheelhouse f r r - 8- temperature measurement also differed between participating countries (Table 2gJ; given the varying·degrees of detail submitted, no unified code has been devised in this case. To allow some quantitative analysis to be made of the exposure of the various sensors in use! an exposure index was devised (Table 2f). TwO' independent assessments were then made of the index to be applied to each sensor on each vessel, depending on the relative wind direction, .using the ship details and photographs supplied. These assessments proved very close in the majority of cases, and a brief discussion achieved, consensus in the remaining cases. The resulting Exposure Indices are shown in Table 3. It can be .seen that five ships have not had exposure indices estimated" due to lack of information1. 2.3 Ship Routes It was intended that vessels recruited to the project should spend a high percentage of their time inilie North Atlantic area. In the event, the recruited vessels followed 4 basic types of route (Table 4), some of which took the vessels outside the analysis area which was defined by the coverage.ofthe lJKM:O Fine.Mesh Model. (30 - 80oN; Soow - 400 E). 2.4 Editing of the VSOP-N'A data The question of numerical errors in the data carmot be ignored; for all observations except those provided by FRG, .the data have' not been subjected to quality control procedures~ and so errors in reported variables and incorrectpositions will be present in the' dataset. There will also be occasions when the model field may not provide an appropriate comparison with observed values for certain variables (for example for ships very close to land). In order that the dataset should be reasonably representative ofreal-time data disseminated on GTS, it was not considered appropriate to conduct exhaustive quality control checks on the data. Isolated observations which probably contained errors in the .reported ship position were deleted from the comparison dataset; such errors may have occurred whilst coding or archiving. Geographical filtering was applied to avoid inappropriate comparisons with the model fields. It was considered that observations made close to the coast should not be used in the final analysis since they could be subject to local effects which the numerical model would be unable to represent. In such cases, a comparison between the observed and model variables would reflect more on the shortcomings of the model than on the quality of the observation. For this reason, the observations from the areas shown in Figure 2.1 were not included in the Comparison Dataset Further, during analysis of the data at the lames RennellCentre for Ocean Circulation (JRC) it became clear that model air temperatures off the east coast of the USA were anomalously low under certain conditions. A further area of data was therefore excluded from analysis at the JRC; Figure 2.2 shows the distribution of data in the fmal analysis set 1 For a detailed description of the instrument exposure for each ship. and of the ship routes. see the VSOP-NA ship catalogue (Kent and TaylorJ 1991). - 9- I Table 3 EXPOSURE INDICES r SCREENIPSYCHRO:METER ANElvIOf\,1ETER l ReI wind direction 315- 45- 135- 225- 315- 45- 135- 225- SHIP 45 135 225 315 45 135 225 315 l caDS Atlantic Link (UK) S 3 4 6 4 - - - - GBSA Author (UK) S 5 5 7 5 9 9 9 9 GBW Geestbay (UK) S 8 9 7 9 - -- - GBVW Geestport (UK) S 4 7 3 7 -- - - f G]MR Geestcape (UK) S 4 9 6 9 - -- - G]MS Geesthaven (UK) S 4 9 6 9 - -- - GXES CGM Provence (UK) S 7 9 9 9 9 9 9 9 G~ Atlantic Conveyor (UK) s 7 3 7 6 7 4 4 7 VRAZ Nickerie (UK) S 4 8 6 8 9 9 , 9 6 VSBV3 CanmarAmbassador CUI{) S 1 9 7 9 6 5 9 9 DDLN Independ Endeavor (FRG) P 4 8 8 8 - - - - DDUC EuroTexas (FRG) P 4 8 8 8 - - - - DHNE Numberg Atlantic (FRG) p 4 8 8 8 - - - - DHRG Alemania Express (FRG) p 4 8 8 8 - - - - DMC Atoanerica Express (FRG) p 4 8 8 8 - -- - DNBR Independ Concept (FRG) P 8 8 8 8 --- - DNJR Independ Pursuit (FRG) p 8 8 8 8 - -- - PCEL AELAmerica (Neth) P 4 8 8 8 - - - - PELT Gulf Speed (Neth) P 4 8 8 8 - - - - PELU Gulf Spirit (Neth) P 4 8 8 8 - - -- PGDG Nedlloyd Kingston (Neth) P 8 8 8 8 - -- - PODS Nedlloyd Kyoto (Neth) P 8 8 8 8 -- - - PGDW Nedll Zeelandia (Neth) P 8 8 8 8 - -- - PGEG NOOll Neerlandia (Neth) P 4 8 8 8 4 4 4 4 H KRGJ Julius Hammer (USA) s 4 8 1 0 - - - - I KR]L Margaret Lykes (USA) s 3 7 4 7 - --- 1 I KRJP SheldonLykes (USA) u ------KRLZ SealandAtlantic (USA) u ------KRPB Sealand Commitmnt (USA) u ------WMLG Delaware Bay (USA) 11 ------WPFZ Adabelle Lykes (USA) s 0 3 4 3 - - - - f WPHZ Charlotte Lykes (USA) u ------WPVF Galveston Bay (USA) S 1 7 0 7 9 9 9 9 WPWH Nedlloyd Hudson (USA) S 1 7 0 7 9 9 9 9 WSDG Lyra Lykes (USA) S 0 3 3 3 9 9 9 9 [ OXTZ2 Mercandian Sun (USA) P 5 8 9 8 - --- 'OYGK2 Mercand Continent (USA) p 5 8 9 8 - -- - DIDA Ariana crn) s 8 5 9 9 9 7 7 8 FNCl CR Libreville (FR) p 4 8 8 8 6 6 8 6 H FNEF Atlantic· Gartier (FR) s 8 7 8 9 9 9 9 6 FNFD EdouardLD (FR) s 8 9 7 7 9 9 6 9 I FNGM La.Carabie (FR) P 4 8 8 8 4 4 4 4 H FNGS La Lafayette (FR) p 4 8 8 8 4 4 4 4 H I' mOY Jean Charcot (FR) S 5 8 8 7 9 6 9 9 r VSBG8 Irving Forest (CAN) S 5 9 8 9 8 4 3 4 r Key..; P- sling psychrorneter S- two.screens (forthis table exposure assessment assumes measurement in r the upwind location) s - ,one screen r u·- unknown method oftemperature.measurement r H- Hand held·anemometer (all otherwind exposure assessments for fixed anemometers) i~ [ ( l r ! - 10- Table 4. Summary of ship routes Description of Vessel's operation No. of ships No. of Obs. used in analysis Ships spending close to 1000/0 of time within the North 25 14.3k Atlantic region of interest. Ships from NW Europe leaving area at southern 13 -6.9k boundary Ships operating across Biscay and around Iberian coast 4 1.3k Ship operating into Mediterranean and also leaving area 1 'O.1k at southemboundary Ships spending 100% of time outside the region 2 0 TOTAlS 45 22.6k The vessel for which these quality control measures had the most impact was FNFD (Edouard LD). Due to the likely topographical effects on the windflow close to the Iberian coast, through the straits of Gibraltar and along the North African coast. about.75%.of this vessel's observations were ·excluded as being not representative of 'open sea' conditions. For all other vessels. numbers excludedwere very much smaller. It must be. stressed that the total number of observations deleted for the above reasons represent a small percentage of those used for the analysis. The bulk of the observations lost were those made south of 300 N (i.e. out of the Fine Mesh Model· area). In total. out of the 33.6k observations archived. 11.Ok observations were not included in the Comparison Dataset which was used for the analysis. This left 22.6k observations archived along with model values and special VSOP information. plus a further 3.6k.observations and matched model values from·OW$ Cumulus. The Comparison Dataset therefore contains 26.438 observations of which 241 952 fell within the area analysed at the ]RC; the calendar ofmatched observations in the Comparison Dataset is shown in Table lb. 2.5 Data Archiving Two basic datasets were asserIlbled·at the archiving centre at Braclmell. The fIrst (the Observation Dataset) is a copy of the· data sent on the tapes from the keying .centre in Germany. refonnatted into a slightly more readablelayout based on the standard 5-digit codes. The fonnat of this dataset is ·shown in Table5a. The seconddataset (the Comparison Dataset) consists of the VSOP observations and all the special VSOP codes. together. .. with matching values of the. variables taken from an .archive of numerical model analyses. The layout of this dataset is shown in Table 5b (codes used within the dataset were described in Table 2)1. I Copies of these datasets have been sent to the World Data centres A and B. The size of the Observation Dataset is about 4 Mbyte. and that of the Comparison Dataset is 4.28 Mbyte; both are in character format. .,~.---\.~.--."--,.~. --_.\iI~~. ~~- ~.. ~-"'---.""._--~~ \ll'tl~---'~-' -_._-..-'....._,.... -.-,-,,,,.,-_.--.... i ,.; \!i ...... • ---r---'\i,-.....·----' -,.,j TABLE 5a: Format of the VSOPObservation 11 Indicator for method of windspeed report 28- 32 Ship Callsign I Cl =estimated inm/s: 2=anemoin m/s: 37.. 39 M,odelWind Direction (degrees) 3=estimated in kts: 4=anemo in kts) 44-45 M:odel Wind Speed (knots) 13.. 14 Indicator (=9N) for Northern Hemisphere 48-51 Model Air Temperature (deg C) 15-17 Latitude in tenths of degree 53·58 Model Mean Sea Level Pressure (mb) 19 Indicator (=V'J for Western Hemisphere) 61- 64 M:odel Sea Surface Temperature (degC) 20-23 Longitude in· tenths of degree 68·70 VSOP Wind direction (degrees) 25 Cloud amount in oktas 75· 76 VSOP Wind Speed (knots) 26-27 Wind direction (tens of degrees) 79- 82 VSOP Air Temperature (deg C) 28-29 Wind speed 84- 89 VSOP Mean Sea Level Pressure (mb) 31 Indicator (= 1· for dry bulb group) 92-95 VSOP Sea Surface Temperature (deg C) 32 Indicator for sign (O=positive, 1=negative) 97- 98 Country Cede (see Table 2a) 33-35 Dry bulb temperature (tenths of degree C) 100-101 Ship's speed (knots) ~ ~ 37 Indicator (=2 for dewpoint group) 102-104 Ship·s course (degrees) I 38 Indicator for sign (O=positive, 1=negative) 106-107 Cargo Hei~1ht (metres) 39-41 Dewpoint temperature (tenths of degree C) 108-110 Height of Reference Level above MSL (tenths of metres) 43 Indicator (=4 for pressure group) 112-112 Method of Sea Surface Temperature measurement (see Table '2c) 44-47 Pressure (tenths of mb. omitting thousand digit) 113-113 Location of Air Temperature measurement (see Table 2g) 49 Indicator (=7 for weather·group) 114-116 Relative Wind Speed (knots) 50..51 Present weather. code 118.. 120 Relative Wind Direction (degrees) 52-53 Past weather code 123-123 Wlethod of'Wind Speed measurement (see Table 2b) 55 Indicator (=0 for sea temperature group) 124-125 Height (metres) above MSL of Wind Speed measurement 56 Indicator for sign (O=positive, l=negative) 127-127 Exposure of Wind Speed measurement (see Table 2f) 57-59 Sea surface temperature (tenths of degree C) 128.. 128 Wlethod of Air Temperature measurement (see Table 2d) 61 Indicator (=2 for wave group) 129-130 Height above MSL of Air Temperature measurement 62 ..63 Period of sea waves (seconds) 131-131 Exposure of Air Temperature measurement (see Table 2f) 64·65 Height of sea waves (half metres) 132-132 Wlethod of Humidity measurement (see Table 2e) 67·73 CaUsign 134-134 Method of Air Pressure measurement (= 0 at present) 75·76 Country code 135-136 Depth below MSL for Sea Temperatiure measurement (metres) 78·79 Ship's speed (knots) 137·137 T(jtal Cloud cover (oktas) 80-82 Ship's course (degrees true) 141-144 VSOP Dew Point Temperature (degrees C) 84-85 Height of cargo above sea level (metres) 148-151 Model Dew Point Temperature (dearees C) 86 ..88 Height of reference level above sea level (tenths of metres) 90· Method of sea surface temperature measurement 91· Location of air temperature measurement 92·94 Relative wind speed (knots) 96 ..98 Relative wind direction (dearees off bow) * Single figure codes 0-9 as defined nationally - 12- 3. DESClUPTION OF THE FORECAST MODEL 3.1 Choice ofthe Model Since the forecast model analysis was used as a comparison standard, the method by which the analysed fields are calculated, and the effect of the VSOP-NA observations upon,them, must first be considered. At the me Meteorological Office (U"IGv10) , there were two archives of numerical model analyses readily available' for use as comparison material for the VSOP data: the UKMO Global model and the UKM:O Lirilited-Area (so-called Fine Mesh) model. Although the former has the advantage of no limiting southern boundary'(and so all ,submitted VSOPNorth Atlantic observations could be compared with model values), the'laner has a analyses archived per day, against 4 for the fonner. Also, the spatial resolution ofthe Fine Mesh model is double,that' of the Global model and so interpolation errors to an observation point could be expected to be smaller. Thus, on balance, it was considered that the fine Mesh model would yield bener comparison data for VSOP purposes. The effect of the' VSOP-NA data on the model could have been minimised by comparing the observations with a forecast field from, say, 6 hours, previously. Unfortunately, not all the required data was available in the archive of IIT+6" forecasts. The analysis field was therefore used, even though it was calculated using the VSOP-NA data. Some comparisons were also made with the "T+6" hour forecasts to check the validity of this approach. 3.2 Calculation of the Analysis fields 3.2.1 Introduction The Fine Mesh model operates on a grid resolution of 0.75° latitude x O..H375° longitude over an area,30 - aOoN, aoow - 400 E. The fields of interest to the VSOP-NA are: Sea Surface Temperature (SST) 10m Wind Speed 1.5m Air Temperature 10m Wind Direction Relative Humidity (RH) MSL Pressure Essentially, each of the above are analysed in a similar manner, by combining observations with a background field to produce a 'best estimate' of conditions on the synoptic scale. For the atmospheric variables, the ,background field is a forecast field derived from a previous datum time. For sea surface temperature, the background field is simply the previous analysis plus the difference between the current 'climate field and the climate field for the previous day. Climate fields, for a particular' date, are derived by linear interpolation in time between the monthly fields, defined from UKMO datasets of past ship observations. It is important to note that the model does not make use of all observations in the analysis of the various fields; a s~arY of the observation types used in the calculation of the model fields of interest can be found in Table 6; For the "SHIP" reports submined by the VSOP-NA ships, only the pressure, wind, and sea surface temperature are assimilated into the model. - 13- f TABLE 6 Data Types Used by the Fine Mesh Model. The VSOP·NA report "ship" t messages, the Cumulus i. aD OceaD Weather Ship (OWS) I Type No. DescriptioD MSLP WiDd ItJI SST TEMP t 1 Bogus 3 3 3 3 3 2 Temperature Sounding 3 3 3 l 3 Ship Temp Sounding 3 3 3 4 Drop-sonde 3 3 3 5 Rocket-sonde 3 3 3 6 Ship Rocket-sonde 3 3 3 l 7 Pilot Balloon 8 Ship Pilot Balloon 3 9 SATEM 3 10 SYNOP 3 11 SYNOPAUTO 3 12 -Ship 3 3 3 13 Ship auto 3 3 3 ( 14 OWS 3 3 3 15 Airep 3 16 Constant level Balloon l 17 Dril'lh,g Buoy 3 3 18 SATOB 3 19 -HER1v1ES I 20 Compressed Code Satern I 21 ASDAR For use in the model. the observations are extracted from the UKlvIO Synoptic Data Bank. which· contains all data received via the _GTS. After extraction. the data are subjected to quality control procedures _which perfonn validity checks- against background field values interpolated to -., the observation point and also -against other observations _in the area. After likely erroneous data t have-been excluded. the rerr1aining observations are incorporated in the analysis using a complex data assimilation scheme during which weights are assigned to -each observation so that the type of observation and the observation- time (relative to the analysis). can be taken into account. A high weighting indicates observations with high reliability taken close to the time of the analysis; such an observation would have a-large influence on the analysis. TIle weighting can be set to zero so any particular observation type can be ignored by the analysis ifso desired. I Bogus observations- (observations createdby forecasters in real time ·and based on qualitative evidence. such as satellite pictures) may have been added to the analysis with a relatively high weighting if it was considered at the time that ·the model was not representing a -feature- well. The f archived analys~may thus contain a small element of subjectively introduced information -at times. f 3.2.2 Sea Surface Ternperature The sea surface temperature (SST) analysis scheme. which is updated once per day. is based on observations from seven different data sources. a climatologicalfield of SST's. and a background field (essentially persistence). The data are extracted from theUKlvIOSynoptic Data Bank at approximately 1200· GMT each day-for _observations between 0001 and_2400 GMT for the f previous day. r fr I - 14- After a series of validity checks have been applied, the SST at a gridpoint is calculated by combining climate and backgrolllld fields with valid observations within a defined immediate area. The seven types of observation and their relative weightings in the analysis scheme are: Ships 1.0 XBTs 1.5 OWS 4.0 Satems 0.0 Drifting buoys 1.0 Satobs 1.0 Fixed buoys 1.5 Progressively more weight is given to the climate field and less to the· backgrolllld field as the timemtervaI between observations and the previous analysis increases. 3.2.3 Mean Sea Level Pressure,·Wind Speed and Wind Direction The calculation of'mean sea level pressure, wind speed and wind direction is more complex than that for sea surface. temperature. since the resulting fields of pressure and wind· must be themselves consistent. and also consistent with analyses· at higher' levels. Whilst sea surface temperatures are updated every 24 hours. analysis of these variables is updated at 3 hourly intervals. As pressure varies reasonably smoothly. on the synoptic space scale and a representative measurement can be obtained relatively easily. pressure observations over both land and sea are used by the analysis scheme. However. surface wind speed and direction are more susceptible to localised effects sllchas shelter and turbulence and so land surface wind reports are not used in the analysis. Wind observations over the sea are generally not subject to such major local effects and so are considered sufficiently representative for inclusion. The data first pass through. the analysis stage which incorporates the automatic quality control procedures and the calculation of weights to refie et the standard errors of each.type of observation. The data are then. introduced to the model during the assimilation stage. during which the data are used to correct the field at each time step of a 3 hour forecast ending at the analysis time (for more detail see Atkins &Woodage. 1985. or Bell & Dickinson.. 1987). Winds from the lowest standardlevel (approx 25m above sea level) are then extrapolated to lOm above sea level. 3.2.4 Temperature and Humidity Like the pressure and wind fields. these variables are updated at .3 hourly intervals. No surface observations of temperature or humidity are used by the model as they are considered to be susceptible to local anomalies and thus may be unrepresentative on the synoptic scale. However. temperature and humidity observations at standard levels from radio-sonde ascents ·are given high weightings by the analysis. The only ships mthe VSOP-NA project which lallllched radiosondes are theCanmar Ambassador. VSBV3 (which carries an ASAP system). and the OWS Cumulus· (which launches 6 hourly radiosonde ascents). Since the frrst standard level is the surface,this implies that surface temperature and humidity data from these ships. but not other VSOP-NA ships, are used by the analysis. I l - 15- I, The data are assimilated into the analysis in the same manner as the pressure and wind t observations. and the surface values are extrapolated from the lowest standard level to provide a I 1.Sm 'surface' value. For ease of comparison with u1.e observations. the model hillrridity values archived in the VSOP-NA dataset are dew point temperatures calculated from the model relative t humidity and dry bulb temperature. I. l 3.3 Model Treatment ofthe Boundary Layer I. Since the model operates using a-coordinates in the vertical ( (J = p/p*. where p* is the local surface pressure). an extrapolation from the lowest a-level (al) is necessary to produce surface variables for direct comparison with observations. A bulk Richardson number is calculated at each gridpoint to determine the most appropriate profile to use to determine the 'surface' values of wind speed (at lOm) and potential temperature (at I.Sm) from the allevel. The profiles used are derived from Monin-0bukhov similarity theory; more -detail is given in Bell & Dickinson (1987) and Devrell l et al (1985). The direction of the surface wind is assumed to be that of the wind at the al level. t r r 4. STATISTICAL SUMMARY OF THE ANALYSIS RESULTS Following the preparation of the Comparison Dataset by the UKNIO the analysis was I conducted jointly at the UKlvfO and the ]RC. Table7J prepared by the UKlvfO. summarises the mean differences (observation minus model) and standard errors -of these means. by variable. by ship and by participating country. The size of the standard errors indicates that. in many instances. differences between and within national fleets are significant. In some casesJ it is I possible to explain differences by simple consideration of some _aspect of instrument exposure. but in others the explanation is not obvious! and further discussion is deferred to sections 6to 9. where r more detailed aspects are discussed for each variable separately. Values discussed below are mean r departures from model values. r 4.1 .Air Temperatures f Over allVSOPobservations the average is close to zero (i.e. -O.07~C ±O.Ol); however. the OWS Cumulus shows an average of -0.61 °C ±0~02J suggesting that the model displays a positive bias of this order~ Only 3 vessels show averages significantly below that of the OWS. suggesting f that the -majority are subject to warming of air within the ship's boundary layer. Both methods of r measurement appear among the highest positive values. suggesting that psychrometersas well as screens can be ~ubject to exposure problems. However a-possible geographical trend in the model r values (Section-7.3) may have biased some of these statistics. r r -4.2 Dew Point Temperatures Over an VSOP-observations the average is +0-.24°C ±0.02J against-an OWS Cumulus average of -0. 15°C ±0.04. This suggests that positive errors -due to inadequate wetting of the -wet f bulb thermometer are common; ships displaying the largest positive values 'all use screens rather r r r r TABLE 7 : (OBS-MODEL) STATISTICS BY SHIP AND COUNTRY AIR 1EMP(C) DEVV POINT(C) SEA TEMP(C) WIND SPEED(kt) WIND DIRIN 'Deg) MSL PRESSURE (mb) Callsign No Mean SE No Mean SE No Mean SE No Mean SE No Mean SE No Mean SE Callsign C6DS 508 -0.67 0.08 351 2.50 0.21 471 0.12 0.08 503 0.82 0.36 494 2.69 1.39 511 -0.20 0.06 C6DS GBSA 455 -0.42 0;06 334 -0.00 0.11 379 0.03 0.06 453 4.03 0.34 449 -1.12 1.50 456 -0.01 0.07 GBSA OBW 999 -0.16 0.06 757 0.40 0.09 981 0.11 0.06 980 2.06 0.17 906 -1.36 0.91 1000 -0.21 0.04 GBW OBVW 944 -0.11 0.03 676 -0.21 0.08 944 -0.40 0.02 923 0.49 0.13 798 -1.38 0.82 944 -0.20 0.03 GBVW OJMR 1184 -0.11 0.05 943 0.03 0.07 961 0.35 0.06 1164 1.87 0.14 1080 -1.18 0.90 1184 -0.15 0.04 GJMR OJMS 868 -0.26 0.04 691 0.31 0.09 663 0.12 0.04 847 1.51 0.18 824 2.17 1.19 868 -0.38 0.04 GJMS OXES 1141 -0.37 0.04 850 0.91 0.09 1139 0.07 0.04 1133 2.29 0.16 1110 1.26 0.88 1141 -0.16 0.04 GXES OZ!v1M 808 -0.91 0.07 573 -0.52 0.11 806 -0.11 0.06 795 3.23 0.24 771 2.59 0.89 809 -0.16 0.07 GZMM VR.AZ 550 0.27 0.06 345 0.33 0.11 548 0.44 0.05 546 2.10 0.25 518 -1.71 1.36 552 0.10 0.05 VRAZ VSBV3 1367 -0.06 0.07 763 0.95 0.09 1362 0.23 0.06 1364 3.16 0.15 1349 -0.79 0.58 1372 0.33 0.03 VSBV3 All UK 8824 -0.31 0.02 8185 0.28 0.03 8254 0.09 0.01 8708 2.21 0.05 8299 -0.25 0.23 8831 -0.05 0.01 All UK DDLN 751 -0.17 0.05 475 -0.50 0.12 739 -0.15 0.07 749 6.47 0.33 726 -2.25 1.24 750 -0.44 0.05 DDLN DDUC 148 -0.15 0.10 69 -0.18 0.28 130 -0.30 0.10 147 7.05 0.81 141 -1.10 2.78 148 -1.18 0.16 DDUC DHNE 974 -0.22 0.06 738 -0.10 0.10 266 -0.16 0.11 973 3.70 0.22 939 0.25 0.99 978 0.19 0.07 DHNE DHRO 772 ~0.60 0.06 557 -0.95 0.11 287 -0.23 0.08 765 2.64 0.22 729 0.25 1.24 772 -0.49 0.06 DHRG DIMC 721 -0.52 0.06 498 -0.62 0.10 403 0.11 0.07 714 3.02 0.25 684 0.79 1.27 721 -1.63 0.11 DIMC DNBR 458 -0.50 0.07 222 0.23 0.16 304 -0.10 0.07 457 2.10 0.28 439 1.03 1.54 459 0.48 0.10 DNBR DNlR 598 -0.39 0.07 463 -0.41 0.10 559 0.24 0.09 590 1.36 0.24 551 -2.54 1.58 565 -0.03 0.13 DNlR All FRO 4422 ~0.37 0.02 3022 -0.43 0.05 2688 -0.04 0.03 4395 3.51 0.11 4209 -0.42 0.51 4393 -0.38 0.04 All FRG PCEL 258 0.45 0.12 124 0.16 0.23 257 1.14 0.17 258 2.31 0.39 256 1.32 1.66 259 -0.02 0.06 PeEL PELT 153 0.38 0.14 64 -0.66 0.28 152 2.34 0.24 153 3.82 0.54 151 -2.11 2.56 152 -0.21 0.09 IllT PELU 175 0.39 0.11 74 -0.46 0.23 175 1.20 0.16 173 4.35 0.50 170 -6.05 2.29 175 0.10 0.59 PELU PODO 472 0.44 0;06 318 -0.04 0.14 470 0.09 0.04 457 0.76 0.24 445 -1.79 1.61 472 -0.40 0.06 PGDG ...... PODS 420 0.79 0.07 173 -0.43 0.15 392 0.38 0.06 397 1.25 0.36 375 -5.42 1.76 421 -0.45 0.07 PGDS Cf PODW 196 0.08 0.08 110 -0.13 0.22 196 -0.21 0.07 191 1.80 0.39 183 -3.29 2.85 196 -0.68 0.10 PGDW POEa 438 0.00 0;06 199 -0.60 0.14 355 -0.34 0.06 425 2.18 0.32 409 1.95 1.60 438 -0.62 0.07 PGEG All Neth 2112 0.38 0.03 1062 -0.26 0.07 1997 0.44 0.04 2054 1.97 0.14 1989 -1.83 0.73 2113 -0;38 0.06 All Neth KRG] 346 0.34 0.09 240 0.01 0.15 348 2.12 0.13 322 1.76 0.36 313 4.50 2.55 349 -0.11 0.08 KRGJ KRJL 539 -0.74 0.10 408 -0.04 0.13 534 -2.37 0.16 525 3.71 0.32 524 -1.48 1.73 539 -1.77 1.69 KR]L KRJP 298 0.24 0.09 145 0.42 0.19 298 0.19 0.12 298 3.52 0.37 297 -5.40 2.32 298 -0.04 0.11 KRJP KRLZ 414 0.91 0.15 283 1.02 0.15 2 7.70 10.75 393 3.73 0.34 388 1.11 1.88 414 -0.77 0.12 I The average over allVSOP observations is +0. 12°C ±O.Ol. against an OWS Cumulus average l of +0.05°C ±0.01. Most of the ·highest positive values come from ships using a water intake method of measurement; only one vessel. again using an intake method. shows a large negative I. value. 4.4 WladSp••ds Over all VSOP observationsJ the average is +3.03knots±O.04., ". against an OWS Cumulus [ average of +2.64 knots ±0.07. No ship generates a negative mean difference from the model. I Each method of measurement gives rise to both high'and low averages. and the standard error also t varies substantially· between ships. r 4.5 WlndDlrectioDs Average difference "overall VSOPships is -O.52°±O.23 J against anOWS Cumulus Clverage of +4.76°±O.29. Vessels with Jarge standard errors for wind speed also tend to displaylarge standard errors for wind direction. f r 4.6 Pressues ~ Average difference over all VSOP ships is-O.29mb ±0.04. against anOWS Cumulus average of +0.10 mb ±O.Ol. This negative bias may reflect the prevalence of inadequate height ( corrections (however·see.section9). Precision'aneroidbarometers seem in general to give rise to smaller mean differences and standard errors. 5. INTRODUCTION·TO THE DETJULED ANALYSIS r Detailed results, for each variable ·,are discussed in' sections 6 to. 9. For each variable the distribution of.model values represented in the data set is fIrst presented to indicate the range of r values for which the results are. likely to be·valid.. The various data comparisons then follow. For a r variable. V, the results are· normally plotted as the mean·difference. dV. binnedagainst.model values. that is: r (1) Ni f where Vo is the value observed by the VSOP-NA ship and VM. is the corresponding modelvalue. r The· summation is over all values of (V0 -- Vw corresponding to VM values in some range VMito (VMi + dVM). N i is the number of observations lying in that· range. Where. the results are plotted for r individual shipsl the order of ships·shown is in alphabetical·order·of·recruiting country..(Canada. f r - 18- France.. Gennany. the Netherlands. UK and USA). and. for a given country. in order of the ship call sign. This is the same order as was used by Kentand Taylor (1991).1. Where error bars are shown on a plot these represent the standard error of the mean. O'm. defmedby: (,~. = [x i- xm ]2)0.5 (2) Gm ~- N (N - 1) where x iand x m are the individual values and the mean value. respectively. N is,the number of values. In some cases care must be taken -in interpreting the plotsJ particularly for the extreme values where there are few observations which may only comefrom one or two ships. In these cases a bias for a particular ship- could result in small error, bars but a ,totally unrepresentativem~an. Also where there is an artificial limit to the values. this will result in a mean bias in the obserVations., 'For example wind speedcannbt be reported as less than zero. so' observations in light winds are bound to be biased high. It is possible that similar. artificial. limits affect the air and sea temperatureS where-a separate'nag'has to be set to report a negative value. The effect of solar radiation (for example on air temperature measurements) has been represented by calculating the shortwave insolation (in W/m2) using the local sun time and the ships' observations of cloud cover. as in ,Smith and Dobson (1985). Night-time values are-,-plotted as a function of the observed cloud cover (in oktas). Because the data set used forqetailedanalysis at the ]RC excluded values near the eastcoast of the USA. the results presented in the following section'may differ slightly in value to the overall statistics calculated by the UKM:O and presented in Table 7; however examination shows that any such differences are not statistically significant. 6. SEA SURFACE TEMPERATURE (SST) 6.1R&1l98,ofSST ValuesEDcoulltered· The range of model SST-values which,correspondedto VSOP-NA observations is shown in Figure6.la. The most likely value lay·between 10°C and 15°C with the histogram skewedtowards higher temperatures. Few observations existed for model values below 5°C or above 25°C. Figure 6.Tb shows the same data but presented for 'each ship·separately. Two shipsJDHNE andVSBV3. operated in predominantly'colder water. but-most-ships contributed to the data set over the range l{J'C to 25°C. 1 Confusion as to the oorrect call'sign for the USA recruited ship.Sealand Atlantic (KRLZ)Jmeans that it was listed out oforder in theCatalogueofVSOP-NAships (Kent and TaylorJ 1991). The lncorrectorder has been retained here for consistency with the catalogue. l - 19- 6.2 Mean SST Differences · Ship by Ship f The mean value of the (observed - model) SST difference. ATs. for each sl1ip is sho\Jvn LT} Figure 6.2. The scatter of differences for each ship is shown in Figure 6.3b while Figure 6.3a shows t the overall histogram of difference values. Most values for ATs lie within ±O.5°C of the model l value, which suggests that for SST the model analysis is a good representation of reality. The mean difference was O.lO°C with a standard deviation of 1.62°C. However the distribution is skewed toward positive.difference, that is a tendency for the ships to report values wanner than the model. f Ships with the wannest ATs values (Figure 6.2) are those using engine intake measurements. In some cases (e.g. GulfSpeed. PELT. ]ulius Hammer. KRGD this appears to be a consistent bias since the scatter of observations (Figure,6.3b) is,small. 6.3 Variations due to Recruiting Country r -The differences by recruiting country evident in Figure 6.2 are mainly caused by the choice I. between using buckets or hull contact sensors (Gennany and the UK) -or using engine intake values r (France. ',the Netherlands and the USA). Biases are particularly evident for some ships recruited by the Netherlands (although these values, may have been affected by the short data sets available). and some USA ships. I 6.4 Variations due'to Instrument Type The variation of ATs according to instrument type is shown in Figure 6.4 both for the observed SST compared to the model values (Figure 6.4a) and forATs (Figure 6.4b). For the main r range of model values the hull sensor values are in close agreement with the model values being about O.2°C cold relative to the model at 5°C and about equal near 25°C. The bucket values are l also close to the model and hull contact values except in the lower SST range where they become [ relatively warm. The engine intake values are relativelywann over the entire range. It has been suggested (e.g. Folland and Hsiung, 1986) that, if uninsulated buckets are used, bucket temperatures will be biased low during periods of large heat flux from the sea to the air. Although most, of the buckets,used by the VSOP-NA ships were insulated, uninsulated buckets may have been used on one or more Netherlands and USA recruited ships (see. for example, Figure 17 of Kent and Taylor, 1991). The values ofATs are plotted in Figure 6.5a as a function of the sum 'of the sensible and latent heat fluxes (FH and FE respectively). These'flux 'values were r calculated using the bulk fonnulae: FH= pcpCH{Ta-Tsm ) (3) r FE= pL CE ( qa -CJsm) (4) r- where Ta, and qa are the observed air temperature' and'specific humidity, Tsm and 1 P = density of air" c p = specific heat at constant pressure, ,L = latent heat of evaporation. f r r - 20- Figure 6.5a shows that, as the sea to air heat flux increased, all the SST measurements became colder relative to the model values Since this may be a feature of the model analysis, the data are replotted in Figure 6.5b using the Hull sensor values as a reference. For heat fluxes in the range 50 W/m2 to 300 W/m2 the bucket values agreed within +O.loC to-0.2°Cwith the hull sensor value, but with a tendency for the bucket values to become colder with increasing heat flux. However the bucket values were significantly warm for low heat fluxes and stable conditions. 'Ihe engine intake values behaved similarly but were .. consistently 0.3°C warm over most of the heat flux range. This engine intake temperature bias is similar to the overall difference found in an extensive comparison of bucket and intake temperatures measured on the same ship by James and Fox (1972). It is not.clear whether the colder engine intake temperatures observed at large heat flux values are significant. Figure 6.6 shows the effect of solar radiation on theSST measurement. All the sensors show wanner values relative to the model under conditions of· greater insolation. Compared to the hull sensor the engine. intake values are consistently warm both day and· night. However, whereas the buckets agree with the hull sensor values during the night, the bucket SST's become, warmer with increasing solar radiation. If real,- this could be because the buckets become heated on the deck and do not cool to the SST'temperature. Alternatively it might be caused by near surface temperature gradients in the ocean. 6.5 Effect of Ship Size The James and Fox (1972) study found that, on the ·larger ships, the engine intake temperatures were warmer with respect to buckets compared to smaller ships. The variation of ~Ts with measurement depth (and hence with ship size) is shown in Figure 6.7. The hull contact sensors··are more consistent than the .engine intake values with the latter possibly showing a trend toward warmer temperature as the depth of measurement increased. Figure 6.8 shows, for each·measurement type, the l\Tgvalue for the 19 ships which were less than 200m in length and the 26 ships which were larger. Few ships contribute to the lower temperatures and the results are probably·only significant above .10°. In that· range, engine intake temperatures for the larger ships show a tendency to become warmer as the sea te,mperature increases. For·smaller ships. and far buckets. there .is .the·caolingtrend with increasing SST. The interpretation of these results is not clear. 6,.6 Summary ··SST The- use of different· SST measurement techniques results in biases in the SST data. Engine intake temperatures are more scattered and exhibit a greater bias (overall mean O.3°C ) compared to ,buckets or hull s,ensors. Bucket .. measurements become relatively warmer for lowheatfiux conditions. when the air is wanner than the sea. ·or for periods·· of greater.solar radiation. Intake temperaturesmcreased with sea temperature on the larger ships. l f - 21- I. i l 7. AIR TEMPERATURE AND IlUMIDITY I t 7.1 Introduction l Because air temperature and humidity were measured on the VSOP-NA ships using dry and wet bulb thennometers exposed either in a screen or in a handheld psychrometerJ the two variables will be treated together. The main measurement errors for the dry bulb temperature are likely to l be the "heat islandJ' effect of the shipJ the direct effect of solar radiation on the .instrument, and insufficient time for a psychrometer. normally kept in the wheelhouse. to reach outside temperature. Most of these errors will increase the observed temperature value. Rarely. a dry ~ bulb may become wet through spray or rain. thus decreasing the observed temperature reading. l however a salty dry bulb should still read true. For the wet bulb the main measurement errors are I inadequate ventilation· or a contaminated or imperfectly wetted wick. Each· will result in a L decreased wet bulb depression and increased·dew point temperature. Thus for both·dTa and dTd [ values which indicate the ship observation being relatively high are more likely to.be·in error than those which show the ships reading low. r ( 7.2· aaltgeofValues·EDco'alttereel Almost all model temperatures were in the range 5°C to 25°C , most values being between 1DOC and 2DoC (Figure 7.1 a). Most ships provided observations over this range (Figure 7.1 b). r Dew point temperatures peaked in the 1DOC to 15°C range (Figure 7.2a) andJ as would be expected,. the histogram was skewed toward lower values compared to the air temperature. The r ( distribution of dew point temperatures on a ship by ship basis (Figure 7.2b) was similar to that for air temperature. Some ships (e.g. Ivurnberg Atiantic, Dm'ffi, CanMar Ambassador, VSBV3) encountered a colderrange·oftemperatures because they were solely operating on.the··'Europe to l CanadaJ' route. I r 7.3 Geographical aDeI Seasonal variations The geographical variation of the (observed minus model) differences for air temperature, f dTa and dew point temperature dTd are shown·inFigure7.3.. Both variables showed·a gradient suggesting that the ship temperatures are colder relative to the model to· the north and west r compared to t~e south and east areas. This variation was most·marked for.dTa in winter (Figure r 7. 4b) suggesting a temperature dependent relationship. For .lower temperatures the ship data appear to be colder relative to the .model. Since this means that for· higher temperatures the ship r data are relatively wannerJ solar heating errors in the .ship data might be suspected. However r Figure 7.4cJd shows the distribution of dTa usingnighttime·data only (where nighttirne was defined by the sun being below the horizon). The northwest .. to southeast ·variation.in the ·sign ·and magnitude of dTa.is also present·in these nighttime dataJ ruling out··solar radiation errors as the f cause.' Because the geographical variation of dTa cuts across many of the shipping routes,.· and also I' because the OWSCumulusshowed similar trends to the VSOP-NA ships (Section 7.5), . r I' r ( [ - 22- the variation is believed to be caused, at least in part, by some feature of the model analysis rather than by errors in the ship observations. These biases also are evident if the AT a values are classified, according to season (Figure 7.4e). In the colder,temperatures of autumn and winter the ship temperatures are lower than the model values (ignoring the lowest values where few observations exist). In spring the colder ship values values are lower and'the warmer values are higher, and in summer the ship values are similar. to, or greater than, the model temperatures. 'Plotted as the mean annual cycle of temperatures (Figure 7.4f}, the colder ship values'iJl winter are clearly evident. The differences over the rest ofthe year appear to be caused by the model lagging the observed temperatures. The cycle of Dew Point Temperature (also shown on Figure 7 .. 4f) shows that the ships re.port higher humidities in winter1but'lower humidities in summer, compared to the model. As 'for air temperature, fuemadeI dew point temperatures lagged the ship observations. Given that spatial, and, seasonal variations exist which are believed, at least in part to be due to the model, it is possible that spurious comparison values will be calculated for ships whose reponing period or area is limite(:i. One such ship is the. OWS Cumulus. This must be considered when examining the results. J.4 Variations hy Ship ·aDd Recruiting Country 1'.4.1 Air Temperature The mean values of ATa and ATd for each ship are shown in Figure 7.5. The ATa values show significant differences which seem to depend more on the recruiting country than the type of instrumentation used. Thus psycbrometer values from German recruited ships! and screen values fromUK ships were all up, to 0.5°C lower than the model values. Psychrometer readings on the Netherlands recruited ships, and screens on the USA ships! were generally wann relative to the model by about 0.5°C. The French and USA ships showed greater,variation than those for other countries, with ,the French ships which used psychrometers giving higher air temperature values than other ships. Because of the possible geographical variation, in the model. there is the possibility that these, differences are. at least in part. due to the ships recruited by different countries 'operating on different routes. Thus the German ships operated on the Europe to No,rth America routes. the Dutch and French ships on the Europe. to the Cape or the Straits of Gibraltar. However. Figure 1. .5c shows the mean air temperature differences based only on data for the region 45°N to SOON. lOoW to 20:°W.. and the period February 1989 to February 1990. This shows similar country to country,variations of !i.Ta to those for the whole data set; for example the Netherlands shipst values arewarmt the USA ships show a large scatter ofvalues. These characteristics are also evident in the ship by ship distribution of ATavalues (Figure 7.6b). Considering the overall distribution of ATa (Figure 1.6a), most values were in the range±1.5°C suggesting thatthe biases between the model and the ships became small when averaged over a large area and period. The mean difference was -0. 16°:C with a standard deviation of 1.73Q C. f [ - 23- I. l 7.4.2 Dew Point Temperature In contrast to L\Ta the mean values for L\Td (Figure 7.5b) appear to be more dependent on It the instrument method than the COlll1try or ship route. The lowest, and therefore probably most t reliable values. were obtained using psychrometers and were up to 0.5°C below the model values. Screen values were generally wanner than the model by up to I.SoC. It is notable that, although l the French psychrometer L\Ta values were biased high. the L\Td values were similar to the screen l values on other French ships. As for ATa• overall, most ATd values were in the range ±1.5°C but t with askew toward positive values (Figure 7.7a). Most ships contributed to this skewness (Figure 7~7b), however the distributions for the German and UK recruited ships were more sharply peaked than those for the Dutch or.USA ships. The mean difference was 0.10°C with a standard deviation of 1.62°C. 7.5 Variations due to Instrument Type tr l The temperature and humidity instruments used by the VSOP~NA ships were either psychrometers, or screens. The French ships .which used. the "Pommar" meteorological'system. r r with platinum resistance thermometers in a small "stackof plates" screen. have been treated separately (labelled "PRT" on Figures 7.8 and 7.9). These were the Edouard LD (FNFD), Atlantic Cc.rtier(FNEF). and the]ean Charcot (FNOY). 'The values forl1Ta are plotted as a function of the model temperature in Figure 7.Bb. Each of the instruments.shows a trend such that the ships values were wanner relative to the model at the higher temperatures. This trend is also shown by the data from the weather ship Cumulus, which f uses screens, suggesting that the··trend isa feature of the model values. On average the ( psychrometer values'are wannest and the PRT values coldest. however the differences are only of order ±O.l °C. The dew point temperature differences, L\Td·(Figure 7.9). show both· a trend compared to the model. and variations between instrument types. For all instrument types. and for the OWS Cumulus. the value of ATd becomes colder relative to the·model with increasing temperatures suggesting that compared to the model the ships observe higher humidity·at low. temperatures and lower humidity at higher temperatures. Thepsychrometerand .PRT values were lower, and therefore'more likely to be correct, compared to the screen values. However the Cumulus values. based on screen readings. were lower still. There. is ·apossibility that this could be partly a geographic effect. however it -suggests that all the VSOP-NAship data overestimate the humidity f values to a greater orlesser.extent. r 7.6 Effect'ofInstrument Exposure r The exposure rating of the· temperature sensor on each ship (Tables 2f and 3) was assessed for winds within ±45° of the bowJon the beam (45 0 to 135° and 225° to 315°) and from ±45°ofthe f stem. The histograms of relative wind speeds and directions (Figure 7.10) demonstrate that, since the. ships were normally travelling at 16 to 18 knots (Kent and Taylof. ·'1991). the relative wind is r predominantly from the bow (690/0. compared to 28% on the beam and 30/0 from astern). The most r important instrument exposure rating is therefore·that for.winds on the bow. The exposure ratings achieved for the VSOP-NAdataset are shown in Figure 7.11. Psychrometer readings were normally r r r - 24- taken on the bridge wing with an exposure rating of 4 or 8. Screen sitings resulted in more varied exposure ratings (see for example the UK ships in Figure 7.11b). On some USA ships the screens were very poorly exposed with low exposure ratings. The variation of the air temperature. difference, ~Ta, with exposure is shown in Figure· 7.12a for good (6- 9), medium (3 - 5), and poor(D - 2) exposure ratings. For the night-time values, plotted against cloud cover, .the ~Ta values show little variation with exposure; .the poor exposure values may be about O.2°C ··warmer, but the difference is hardly significant. Day-time values are plotted against the solar radiation (see section5.2J. The ~Ta value becomes more positive (ship temperatures wanner) with increasing solar radiation. The effect is greatest for poorly exposed sensors, with medium· and well·exposed sensors showing relatively smaller biases. The effect of solar radiation does not seem to depend greatly on whether screens or psychrometers are used (Figure 7.12b). Since at night the psychrometer values are about O. 2°C warmer, and since in the mean· the psychrometers are warmer.than the screens, it may be that th~ screens show a greaterradiative effect bYi at most, about O.2°C to O.3°C. There isa marked dependence on the .relative .wind speed. Stronger valuesof·the relative· wind decrease ·the solar radiation effect considerably (Figure ··7 .12c). The ~Td values ·also show exposure related effects (Figure 7.13a). The reason ·why the poorly exposed sensors·should read high lll1der increased solar radiation is not clear, since the dew point temperature should not be altered by .sunshine effects. The medium and well exposed sensors both show a different behaviour, ·with the ship values relatively dry for conditions of low cloud cover (whether by· day or by night). The effect occurs for both psychrometer and screen values (Figure 7.13b) although it is more marked in the latter. This suggests that the model overestimates the humidity when· the· sky is relatively cloud free,. and that lll1der cloudy conditions (when the relative humidity is likely to be higher), the screens overestimate the humidity relative to the psychrometers. 7.7 Summary· Air Temperature and Humidity Geographic trends of order I°C were detected, particularly in the air temperature comparisons, which are thought to have been due to variations in the model values. The results also suggested significant mean biases in the model values for temperature (biased lower for low temperatures). Psychrometer .measurements of air temperature were similar to J or slightly wanner (by about 0.1 °0 ) compared to those from screens, but the psychrometer humidity values were significantly lower by as much as 10C (and presumably more accurate) than most of the screen values. However the OWS Cumulus, which uses screens, returned dew point temperatures even lower than the psychrometer readings suggesting that all the VSOP ships overestimated the atmospheric humidity. In general the exposure of psychrometerswas better than that for screens. The observed air temperature values were higher for poorly exposed sensors (screens and psychrometers) due to day time solar heating. This effect decreased with increased wind speed. Under cloudy conditions (when relative humidity was presumably higher). the screen values of dew point were relatively high by about 0.9°e compared topsychrometers. There was also evidence that the model overestimates the relative humidity for periods of light cloud·cover. - 25- l There were variations which depended on the country which recruited the ship. The l "Pomar" type screens on the French ships gave humidity values more nearly comparable to the I psychrometer readings. Psychrometer air temperatures on the Dutch sr.aips tended to be \Jl{a...Y'Jn by about 0.75°C. Instruments on some of the USA recruited ships had poor exposure. and the t temperature values from the USA ships showed large biases and scatter. f ( 8. WIND OBSERVATIONS 8.1 Range ofWind Value. Encountered I The .distributioIl ofmodel wind speed values·corresponding to .the VSOP-NA observations (Figure 8.1 a) peaks in the range i0 to 15 knots (5 to 1.5 mls) 1. the· most.likely values were in the f range 5 to 25 knots.with almost all ships reporting wind speed values in thatrange (Figure8.lb). The most likely. wind .direction was about 225° with. ·as might be expected. significantly more ( observations ofwinds from directions from the west thandireetions from the east (Figure B.le). [ 8.2 GeographicalVarlatioDs The ship'minus model wind speed differences. 4vv. were positive over the. whole area. that is comparedto·the model the ships reported.higher·winds. There was·a·geographicaltrend with I larger values •. to the .northand.west compared to the south andeast.(Figure.8.2) however this was ~ probably related to the.distribution ofwind·speed values (Figure· 8.3). 8.3 Wind Speed andDlI'ection- Ship by Ship t 8.3.1 Wind speed The mean value of the ship minus model difference for wind speed. aw, and direction, Addd. are shown in .Figure 8.4. The VSOP-NA ships used fixed .anemometers. hand-held anemometers. or visual (Beaufort) estimates to.deterrninethe wind. The,.Avv values-were positive for all· ships, no matter what instrument type was used. The overalldistributionofL1vv. was also 'skewed to positive values (ships reporting higher), the most likely difference being about 1.5 knots compared tolhe mean difference of 2.9 knots. The standard deviation was about 5 knots. This f skewness·is evident· in the 'individual ship values ofAw (Figure a.5b). It·might be caused, at least r in part, bymiscoding or transmission errors which can impose large positive,. but not large r negative. wind speed errors. Hall et al. (1991) compared VOSdatato theal (25m) model wind value. fortheUK1v10 model for the periodJanuary toJune 1990. ,For data which had been quality controlled they found, f a near Gaussiandistributionfor L\vv with· a bias of 2.4 knots and a standard deviation of 6.6 knots. r 1 Ships report winds in either knots ormls.To simplify comparisons with the ship speeds. this r report uses knots. To the accuracy required here,. 1 knot = .0.5m/s. r rI r I - 26- On the basis of a mean difference of only-O.2 knots between the model and OWS Cumulus, they concluded that the vas ships' observations ,were too large rather than the model winds being too low. However in these data the OWS CumulusAvv value, corrected to lOm height. is about 2.6 knots which is a similar to the VSOP-NA mean difference. Thtisr ·overall1 these results·suggest that the model under-estimates the wind-speed by about 2 knots. The wind reports from the VSOP-NA ships~are used~'bythemodel in calculating the analysis field. However·comparisons using the 6 hour forecast fieldsfrom.the •. modelwere·not significantly different t.o those based.·.on the·analysis ·results. 8.3.2 Wind Direction The mean.Addd values for almost an'ships (Figure 8r4b) were within ±SO of the IIlodel value. Six of the ten VSOP-NA ships· which used wind vanes were among the 7 ships with mean differences greater than ±5°. Even the OWS Cumulus, which uses wind vanes, showed a bias compared to the model of 5°. This probably·illustrates the difficulty of aligning a wind vane to the ship's head to better than ±lO°. Most Addd values, whether visual or instrument derived, Were within ±10° of the model value (Figure a.5c). Outside this range, the scatter of Addd values was similar for the two methods with most ships observing wind values up to 90° from the ·model value and sometimes more (Figure a.5d). Some of these differences may have been due to small scale features ofthe wind field which are not represented in the model. In other cases coding'errors may have occurred (see. Section 8.6.4). 8.4 VariatioR with Wind Velocity The variation ofAw with wind speed is shown in Figure 8.6. The positive bias ofthe ship observations compared to the model is evident at all wind speeds. . In the most commonly observed range, 5 to 25 knots, the difference increases with wind speed. Both the .. model and the ships showed. a variation of the mean wind speed with wind direction with winds from the ·west being strongest (Figure a.6c). This causes an apparent variation of Aw with wind direction (Figure 8.6d)., 8.5 Variation due to Method of EstimatioD The variation of Avvwith wind speed was different depending on the estimation method used ·(Figure.8.7aJb). Visual winds (which dominated in the overall mean relationship shown in Figure 8.Bb) were lower than winds from fIXed anemometers. The difference was about 2.5 knots for the commonly observed wind .speed range. Below 15 knots. winds from hand-held ane'IDometers gave similar Avv·values to visual winds. At higher·wind speeds the few observations which were obtained showed large scatter. These differences were not related to theatm.ospheric stability (Figure 8.7c). Stablecases were defined as the air temperature being'wanner than the sea. Under stable·conditions visual winds'may nave been estimated slightly higher at low wind speeds·but there was little significant difference. - 27- l l 8.6 Anemometer Wind Data I 8.6.1 Variation with Recruiting Country t The fixed anemometers on French and USA recruited ships (Figure 8.8) showed similar behaviour. both being significantly different to the OWS Cumulus anemometer values. This was l true whether the model analysis results or the forecast values were used for the Cumulus comparison. t l. 8.6.2 Effect of Anemometer Height For those ships using anemometers. dvv showed a possible dependence on the· anemometer height (Figure 8.9). Also shown on this plot is the required correction to the wind speed to give 10m values assuming neutral·stability·(Dobson, 1981). Clearly this correction does not explain the difference between the anemometer readings and the model. Even allowing for a model offset, the actual dependence on anemometer height appears to be greater than would be expected from the f variation of wind with height. Possibly this is due to increased acceleration of the wind flow over larger ships. f r 8.6.3 Variation with Instrument Exposure The exposure rating of the anemometer for those ships carrying fixed anemometers was defined in Tables 2!- and 3. In general the anemometers were well exposed with good exposure ratings being achieved both overall (Figure 8.1 Oa). and on an individual ship basis (Figure 8. lOb). )- Most observations (880/0) were categorised as having a good exposure rating (6 - 9), relatively few (120/0) fell into the medium category (3 - 5), and none were considered poor (0 - 2). In terms of &vv. there was little statistically significant difference between the good and medium exposure categories (Figure 8.11); the medium exposures perhaps gave lower winds in the 15 to 25 knots range. ( 8.6;4 Errors in estimating the. True Wind The VOS ·are requested to report the true wind speed and direction. For those ships using anemometers. .the ships' officers must perform the vector subtraction ofthe ships velocity from the measured relative wind. Given that the typical speed of the VSOP-NA ships was 16 to 18 knotsJ and r that the most likely wind speed was in the range 5. to 20 knots. a·large error can result if this r calculation ig·not performed correctly. To detennine·thelikelihood of this error the VSOP-NA ships were requested to report ships speed and head. and the relative wind speed.and directionJ in r addition to the true wind values. Since most VSOP-NA wind data were visual observations. the r number of these relative wind. observations reported was small. (about 2500). Using these data, the .. relative wind has been calculated from the reported true windJ and compared to the. observed , relative wind. Figure 8.l2a .shows that only about 50% of the reported winds· corresponded· to f calculated relative winds within ±2 knots of the observed value. A large. fraction of the· reports r (300/0) were more than±5 knots different. For wind direction only 10% were within ±lO°. and 130/0 were outside ±50°. These are large errors which significantly degrade the dataset of r anemometer winds. r r ~ r - 28- 8.7 Visual Wind Data 8.7.1 Variation with Recruiting Country For ships· recruited by the Netherlands, UK, and USA, the variation of~vvwith.wind speed for visually estimated winds was .very similar (Figure 8.13). The German ships produced significantly higher wind values, although· these were less than the anemometer derived values discussed. above. 8.7'.2 Comparison of Day-time and Night-time values. It might be expected that· visual estimation of the wind is more difficult at nightthan during the day. Figure· 8.14a shows the anemometer and visual winds for day~timeandnight-timeperiQds where night was defined as the sun being below the horizon. The ~vv values· for the anemometer winds are not significantly differentbetween the two periods. However with'-increasing wind speed the Beaufort estimates ·atnight become10wer·than the day-time values. This suggests that at night the ships' officers underestimate the roughness of the sea and hence the wind speed. Six of theVSOP-NA ships which returned visual wind reports were also equipped with anemometers. Figure8.14b shows that, for day-time observations, the ~vvvalues for these ships were no different to those from other ships returning visual estimates. However at night the ~vv values were similar to thoseobtamed during the day. This suggests that, although the .officers on these ships do not rely primarily on the anemometers for wind estimation, the anemometers are used to help estimate the winds at night. 8.7.3 Beaufort Scale Effects The discrete nature of the Beaufort ·Scale results in a histograrnof visual wind speeds showing peaks at the mid values of each Beaufort Force (Figure a.15a). This effect, which has been noted before (e.g. Wilkerson, 1986), was noted in the wind reports for most ships (Figure8.15b). Over many reports the effect should average out. These figures show' that few ships report wind force 1. A.number·of definitions of the Beaufort.Scale are in use. That recommended for use on the vas ships is knOWIl as HCode lIDO"; a different defmition has. been recommended by the Commission for Marine Meteorology as more accurate than Code lIDO, butthe advantages were not considered sufficientto warrantthe introduction of a new code on the VOS (WMO 1970). A further, rec-ent defmition has been ·suggested by Kaufeld (1981). The effect of using these Beaufort Scale definitions is shown in Figure 8.l6a. The visual winds are increased at most wind speeds and decreased- at the highest wind speeds. Formos!wind speed observations the difference from the model values is increased. The new· scales were dermed by comparing visual· and anemometer estimates. The difference between the visual winds and the anemometer measurements is shown in Figure 8.16b for each ofthe Beaufort Code definitions. None of the codes. gives good agreement over the whole wind speed range; the new codes produce generally better agreement with the anemometer values than Code 1100 but this does not necessarily mean a closer agreement to true conditions. l f - 29- I r 8.8 Summazy· Wind I The model appears to under-estimate the wind, probably by about 2 knots, compared to the observations. Anemometer estimated winds were about 2.5 knots higher than visual winds. t Most anemometers were well exposed and there was little correlation of wind speed differences with exposure. There was a possible correlation between the height of the anemometer and the wind t speed "over-estimate" with high anemometers giving stronger winds than would be expected from the ,variation of wind speed with height. Wind vanes were prone to fixed directional errors which f were not present in the Beaufort estimates. The conversion of anemometer measured relative wind velocities, to true wind introduced many significant errors into the dataset. [ For winds above about 15 knots. visual wind reports underestimated the wind at night compared to day time periods. On those ships reporting visual winds which were also fitted with anemometers. the night-time winds were similar to the day-time visual winds suggesting that the anemometer is being used to help. but not replace'the visual'estimate. Use of alternative Beaufort I ) Scale definitions produced better agreement between the visual'andanemometerwinds. but worse I. agreement with ,the model. No single scale produced good ,agreement, over the entire wind speed range. r [ 9. PRESSURE 9.1 RaDge ofPressure Values Encountered l Most ofthe observations corresponded to model pressure values between 1010 mb and 1030 mb (Figure 9.la) with most ships contributing data in the range 990 mb to 1040 mb (1 mb = IhPa). r ~ 9.2 Geographical Variations The (ship -model) pressure differences AP show, a significant geographical variation with r~ the ships reporting higher pressures with respect to the model to the north., ,.and lower pressures to the south; the, variation being of order 0.3 mb. This variation is apparent in' Figure 9.3 as a tendency to more negative values, ofAP with increasing values of the air temperature. If this variation, was caused. by neglecting the effect of temperature in correcting the ships ,pressure observations to the surface value. then the ships should read higher for- higher temperature conditions; The opposite trend to that in Figure 9.3. Pending further investigation, the geographical variation of AP would appear to be due to themadel rather than the ship observations. 9.3 Mean Pressure Differences - by Ship, Country, and Instrument type The mean values, of APfor each ship are shown in Figure 9.4. AP for the majority of ships was ,between 0.0 and -O.S'mh. ,The'Dutch and'UK ships,used digital Precision Aneroid Barometers (pAB's) and generally, showed less scatterthan'ships recruited by the other countries which used analogue aneroid barometers. Large (order ,I mb) but'consistent biases occurred in the repo1)S . from some of the Gennan and French ships. - 30- The overall histogram of L\P values (Figure 9.5a) shows that most reports were within ± 0.5 mb of the model value, the mean difference was -0.191 mb with a standard deviation of 1.96 mb. Examination of th-e scatter of L\P values (Figure 9.5b) shows that for most ships there were some values which were probably incorrect by 10 mb. Such values would be easily removed by a quality control procedure which would reduce the standard deviation, probably to the 1.6mb found by Hall et al. (1991). Whereas the L\P values for many of the ships lay close to the 'overall mean of-0.2 mb, the OWS Cumulus value was <+0.1 mb. Given the northerly position of Cumulus, this fits the geographical trend in L\P noted above. 9.4 Variation with Mean Pressure Figure 9.6 shows the variation of L\P with the modelpressure value. In the most commonly observed pressure range, 990 rob to.l030 mb, the value ofAI' decreases as the pressure increases. That is. when the pressure is high, the ships. observe a lower pressure than the model. This variation could be associated with the observed geographical variation of dP. A similar variation is not detected in the data forOWS CumUlus, which stays in the same position, suggesting that the trend is geographical or temperature related rather than a function of the absolute pressure. Since'pressure is assimilated into the model during the analysis, it might be expected that the model values and the ship values should be in close agreement. The degree to which the ship values influence the analysis can be' examined by comparing the ship data to the forecast from 6 hours previously. Figure 9.6c shows that compared to the "T+6" forecast the ship values read low by about 0.5 mbmore than.the .comparison with the analysis would suggest. This contrasts with the values for OWS Cumulus which are higher when compared to the forecast (Figure 9.6d). ·This suggests that the vas pressures act to decrease the model pressure over much of the region, whereas the OWS Cumulus data acts to increase the model pressure. A possible cause of this would be inadequate corrections for barometer height on the vas ships. This will now be considered in more detail. 9.5···. Barometer Height Corrections On most ofthe VSO·P-NA ships the barometer was situated atbetween 20m to 30m.above sea leveL At l say, l5°C this means that the barometer will read low by between 2 mb to.3.5 mb compared to the sea surface pressure. To correct· for this, the vas are expected.to. apply a correction factor to the observation. Figure 9.7a shows, 'for each ship, the mean value of L\P plotted against the height. of the temperature sensor. (for most ships this will also be the height of the barometer, however on some ships.the barometer will· be about 3m to 4m higher in which case the ploned point would move higher by up to 0.5 mb). If the ships did not apply a barometer correction the points should lie along the line shown; it would appear that a barometer correction is being applied on all the VSOP-NA ships. This conclusion still holds if the "T+6" forecast data is used instead of the.model analysis data (Figure 9. 7b). The barometer correction to be applied should vary as the draft of the ship varies with cargo load etc. The VSap-NA ships were requested to provide a value for a "reference height" with each observation. Not all ships reported this variable reliably. For those which didJFigure 9.7c shows the value ofAP (adjusted by the mean pressure bias' for that ship) plotted against the variation in - 31- t reference height. Ships represented by points on the zero line have corrected for changes in draft, t ships with points on the sloping line have not. It appears that most ships corrected for changes in / the draft wrJch were over 1 metre in magnitude. SI.uaIler changes appear not to have been allowed t for. f t 9.6 Summary· Pressure Pressure observations from ships using digital Precision Aneroid Barometers appeared to be more accurate than. those using analogue instruments. A few instruments had a consistent I calibration bias of order 1 mb. ( A major cause of mean differences between the ship observations of pressure and the model [ appeared to be a north to south variation in the model values of order O.3·mb. In the mean the ship pressures were 0.2 mb lower than the model analysis. Compared to .the forecast the ship pressures were lower by about 0.7 mb.These differences were greater under conditions of higher pressure.. The lower pressures observed by the ships did not appear to be explained by a failure to f correct for·the.height of, or changes· in the height of, the barometer. r r 10. ·SUMMDY 10.1 Model Characteristics I During the analysis it became clear that some features of the (ship minus model) differences [ were probably caused by characteristics of the model analysis. Furthennore that these trends were such that. for example, incorrect conclusions might be drawn concerning the performance of I ships which operated in limited geographical areas, or for only part of the project. These features r are Slli'TiCIlarised in Table 10. 1. Table 10.1 Characteristics which can probably be ascribed to the model and which must l be allowed for in the analysis~ Variable Characteristic [ ssr Model field appears to be a .good representation of reality. Air Temperature Model values are warmer·under colderconditions I and colder under warm conditions. This leads to a geographical trend; in winter the model values f are wannto the northwest compared to the southeast by about 1.5°C. Model r values lagthe seasonal values.of observed temperature. r HUmidity Model values are relatively drier under lowerhumidities but moister for the most often observed humidities. Model values lag theseasonal·values of r observed humidity. r Winds Model values are biased low compared to the ship data (estimated mean bias =2 knots). Pressures North-south trend in the model values of about 0.3 ·mb~ f r r r r r - 32- 10.2 Instrument Characteristics The VSOP-NA results suggest that there are biases and other differences in the data which depend on the measurement method used. These are summarised in Table 10.2. Table 10.2 Features ascribed to.differing methods of observatioD. Variable Method Characteristic Hull Sensor Consistently good. Engine Intake Greater scatter, some large consistent biases. In mean, biased high by about a.3°C. Bucket Relatively wann (up· to a.5°C ) for low heat flux, stable ·conditions, high solar radiation. Otherwiseagreed with hull sensor "alues. Air·Temperature ~Psychrometer Exposures generally better tban screens. Dutch ships' values high (by about 0.75°C ?). Mean solar radiation error of up to 1.5°C which varies with wind speed. Screen Some poorly exposed (particularly USA). Solar radiation error same as psychrometers except poorly exposedscreens were worse. Dew Point Psychrometer Values lower than screens by 0.5 to 1.5°C. Values low relative to model particularly·under clear sky conditions. Screen Values higher thanpsychrometers particularly under cloudy (higher relative humidity?) conditions (up to 1°C). Screens on UKships showleasrscatter. Winds Fixed Anemo. Speeds were about 2.5 knots greater than visual estimates. Possible dependence on anemometer height even if corrected to. lam. Directions likely to have constant offset. Many errors in conversion to true wind value. Hand Anemo. Speed estimates scattered at wind speeds above 15 knots. Larger scatter for wind direction than other methods. Visual Speeds lower than anemometers. Underestimated for winds above 15 knots at night {unless anemometer on ship). None of Beaufort scale definitions gives perfect agreement with anemometers. Directions generally good. Pressures Digital PAB Digital readings less scattered than analogue. Height corrections seem to be applied properly. Analogue Baro. More scattered, some have offsets of order I mb. l I. t - 33- t ,. 10.3 Augmented documentation A major feature of the VSOP-NA project was the availability of extra information concerning t the participating ships. In some cases this information is available from the "List of Selected Ships" l (WMO 1990) but in incomplete or inaccurate fonn. A particular instance is the presence of anemometers on some ships which return visual wind observations. In other cases, such as the type or exposure of some thennometer screens, the information is not available. These examples f. have been chosen because differences were detectable in the data due to these causes. The usefulness ofthe extra'codes'reported by the VSOP-NA ships is'assessed in Table "10.3. Table 10.3 Value ofthe Extra Code. reponed by theVSOP-NJl ship. rf Extra Information Assessment l _Ship's speed and heading Needed to conv'ert relative 'hind observations. The',speed,and f course made good is reported and may be sufficient. but this r has not been tested~ Height ofdeckcargo Not found useful. Height ofreference level Could be useful,for real time quality control ofpressure data; however. misunderstandings caused this variable to be reported unreliably by some ships. r Method ofSST measurement. There is a definite bias between different methods: some ( ships use more than one method; the'method should be ~ reported with each observation. Location of air temp. Not found useful on observation by observation basis. lr an measmement. Needed as part-of an overall ship desCription. f Relative'wind speed 'and For ships withanemometers~ the vas datasetwould be direction, significantly improved'ifthis was reported instead of or,111 additionto true wind. providedanadequateship's:,speedand f heading'were avallable. r ( f r r r - 34- 11. RECOMMENDATIONS A : RecommendatioDs To Members Operating VOS Al Observing Practices and·Equipment The results of VSOP-:-NA demonstrate clearly the value of national observing fleets conforming to recognised standards of instrument exposure and observing practice. Additionally, for some variables. one method of measurement has been shown to be superior to others ( egSST by hull-contact sensor). For other variables, different methods have both advantages and disadvantages; good exposure is often more important than 'choice of instrument type (see.Table 10.2 in this report). Itis therefore strongly recommended that Members take note of these findings, and ensure that equipment, exposures 'and" observing practices are chosen and maintained appropriately, with. a view to achieving greater accuracy and consistency across the international VQS. Details of hull sensors used by the·Netherlands, .and the UK are included in Kent and Taylor 1991. A2 Publicity for VSO'P-NA Results It is essential.that the results of VSOP-NA be published and made available to all Members. fair relevant dis·tribution"within their countries. Publications 'such as Marine. Observer. Mariners We.ather Log. Wetterlotse 1 Met·Mar and MIM'should be used to·bring 'the results in appropriate fonn to.the attention of theVO:S. Other methods of publicising the project nationally may also be· comidered A3 Real-Time, Data Monitoring The; existing real-time monitoring systems"for VOS~ repo,rts should be extended to cover an variables required for surface flux calculations; specifically vas databases· maintained at eac;p monitoring centre shouldmcludemore detail for eacb ship. to facilitate identification of-the appropriate,·corrections.Results of·the real-time monitoring should be made· available more ft~quently to.Members, and -PMOlS, ideally monthly. A4 Reduction in Reponing Errors TIle results; of VSOP-NA show that many errors were made in converting measured relative wind into true windl and in deriving 'dewpo,int from dry-bulb and wet-buIbtemperatures. Members are recommended to provide Uleir VO:Switb dedicated calculators or co.rnputer pro'grams for deriving: these quantities.., in order to achieve a significant decreas:e in· the number of such erro1S~ [ - 35- f I, l B: Recommendations to WMO I l Bl Sea Temperature Measurements Given the differences resulting from different methods of sea temperature measurement, it I. is desirable that ships' GTS reports should include an indicator to define the method in use. I. However in view of the difficulty of devising a suitable code consistent with the logbook (IMMT) I. procedures and also easily incorporable into the OSnTwTwTw group of the SHIP code, it-is recommended that, before any further action is taken, users concerned with the real-time analysis of SST data should be consulted. in the light of theVSOP-NA resultsJ regarding the value to them of this additional information~ B2 Ship's Draft I: Considering the importance of ship's draft/"heightof eye" reports for the accurate calculation of sea level pressure, and recognising that this -information is at present available only r from logbooks, it is recommended that real-time monitoring centres be consulte-d regarding the [ value of the inclusion of such reports within·each.GTS observation. B3 Code Changes f If a real-time requirement for "method of SST" or "height ofeyeuinformation is found to exist (recommendations Bl and -B2),..then it is recommended that CJvnvI take .action to develop and [ implement appropriate code changes to. effect real-time reporting of this information. r B4 Catalogue of Reporting Ships It is strongly recommended thatWMO improve the accuracy and dissemination of the International ListofSeleeted, Supplementary and Auxiliary Ships (yVMO No. 47). This could be achieved by requesting Members to update and· submit information of a quarterly (rather than an annual) basis, and making the publication available on diskette at similar intervals. Urgent progress.should also be made towards providing the information to· Members through an on-line database. As a longer-term goal, the. expansion of the list to include further· information on types r and locations of sensors should be studied. In this· context, the value of CD-ROM as a medium for r tne storage of drawings, photographs and sketches of ships should be .considered. The VSOP-NA I catalogue·(ret) might be seen as a model for this type of presentation. r BB Observing Practices With the aim of standardising practices -between national observing fleets, it is f recommended that a reference booklet be./prepared to provide concise step-by-step guidance on r observing procedures to vessels of the VQS. Material-_in -the Guide to Marine Meteorological r Services (YVMO No 471) could be used as a starting point. f r - 36- B6 Port Meteorological Officer System The results of the VSOP-NA study demonstrate that an efficient Port Meteorological Officer system can have significant impact on the overall, quality of data submitted by individual" national fleets. It is recommended that appropriate funding and resources be made available to improve the, organization, training and operation of the Port Meteorological Officer 'systems of Member countries. Members with existing, well-established and effectivePMOsystems should be encouraged to to ,'offer training and assistance' facilities' to -other Members to enable them to upgrade their own PMOservices. B7 Beaufort Scale The results of the VSOP-NA analysis of wind data suggest that further consideration be given to the wind speed equivalents of the Beaufort scale. It is recommended tha~ CMlvt, initiatelco~ ordinate further action in this respect, and the' results of VSOP-NA be considered when any proposals are fonnulated. B8 Delayed-mode Anemometer Win~ Recognising_ that the results of VSOP-NA have demonstrated that substantial numbers of errors occur- in the calculation and reporting of true wind speed and direction, both in,real-time (see-also recommendation A4) and in delayed mode, it is recommended that CIVIM re-examine the procedures for 'delayed-mode reporting of anemometer winds,' with, a-view to reducmg'these errors mthe archived data through inclusion of relative wind speed and direction in the logbook reports. C : R.ecommendations for Climate Research Cl Noting -that model-derived ocean surface flux values will be increasingly used for forcing ocean models, and--recognising ,that the VSOP-NA project has shoWn that biases -exist in model derived data such that significant errors would exist in the predicted flux values, it is recommended that increased use be made of the VOS ship observation to verify model flux determinations. C2 It is recommended that where VOSobservations are used to construct sea surface temperature data sets, the observation should be,classified according to measurement type and that greatestweight should be given to I} hull contact sensors, 2) bucket measurements, 3) condenser or engine intake instrumentsj in that order. In particular it should be noted that there is evidence that intake measurements are of poorer quality and likely to be biased warm compared to the other methods (see also recommendations Al 'and Bl). C3 Recognising that ships· observation transmitted over theGTS atpresent contain a significant number of errors due to the incorrect calculation of true wind velocity and dew point, and that these errors-can be reduced by the use of logbook data, the use of delayed-mode'logbook-derived data for climate research is recommended (see also recommendation BB). t f - 37- I. t CA Noting that the greatest accuracy requirements for VOS data are for the calculation of flux I fields for climate research, and recognising that the VSOP-NA project has demonstrated that the i~ quality of ships' data depends on the efficiency of the PMO system, it is recommended that the l climate research community supports measures designed to improve the PMO system. 12. ACKNOWLEDGEMENTS l We would like to acknowledge .the following without whose contribution this report could not have been assembled: the Owners, Masters and Officers of the VSOP-NA ships;. the Port Meteorological Officers, members of the VSOP-NA Management Committee, and staff of the Atmospheric Environment Service (Canada), Direction de la Meteorologie Nationale (France), I Deutscher Wetterdienst Seewetteramt (Gennany), Royal Netherlands Meteorological Institute (mM!, t Netherlands), Meteorological Office (UK), National Weather Service INOAA (USA); and the World j' Weather Watch and World Climate Research Programme (WMO) and the the Committee on Climate Changes and theOcea..., (IOC/SCOR). r The work at the James Rennell Centre for Ocean Circulation .. was partially funded by the r Ministry ofAgriculture, Fisheries.and Food. I 13. REFERENCES f Atkins,M J and Woodage,M J (1985) Observations and Data Assimilation; Met Mag 114, pp.227-233. ( Bell,R S (1985) UKtvIO Operational Numerical Weather Prediction Scheme Documentation Paper No. I 3 - The Data Assimilation Scheme, (Version 2), Meteorological Office, Bracknell. Bell,R S and Dickinson,A (1987) The Meteorological Office operational numerical weather prediction ~ scheme; Met 0 SciPaper No 41, Meteorological Office, Bracknell. Devrell, C, Watkins,F and Kitchen, J (1985) lJKrv10 Operational Numerical Weather Prediction r Scheme Documentation Paper No. 7. Meteorological Office, Bracknell. Dobson, F.W. (1981) Review of Reference Height for and Averaging Time of Surface Wind Measurements at Sea, Marine Meteorology and Related Oceanographic Activities, Report f No. 3, WMO, Geneva, 64pp. Hall. C.D., Ashcroft, J. and Wright, J.D. (1991) The use of output from a numerical model to monitor the quality of marine surface observations, Met. Mag. 120, pp.137-149. James R.W. and Fox P.T. (1972) Comparative sea surface temperature measurements. Reports on Marine Science Affairs (5), (WM0336), WMO, Geneva. r Kent E.C. and Taylor. P.K. (1991) Ships Observing Marine Climate, A Catalogue of the Voluntary ObservingShips Participating in the VSOP-NA, Marine Meteorology and Related r Oceanographic Activities, Report No. 25, WMO Geneva. Folland. C.K. and Hsiung. J. (1986) Corrections of seasonally varying:biases in uninsulated bucket sea surface temperature data using a physical model, Met.O.13 Branch Memorandum f No. 154. Meteorological Office, Bracknell, UK. r Kaufeld L (1981) The development of a new Beaufort equivalent scale, Meteorol.Rdsch.34, 17-23. r r r r - 38~ Smith, S.D. and DobsonF. (1985) Estimation of solar radiation at sea, pp.525-533 in The ocean surface: wave breaking, turbulent mixing and radio probing, (ed. Y.Toba & H.Mitsuyasu). Dordrecht Reidel Publishing Co. 586pp. Starkey, J.P. (1989) VSOP-NA First Interim Report-Data Collection Phase, Interim Report, November 1989" Meteorological Office, Braeknell, UK. Stubbs, M.W. {1981} New eode for reporting surface observations ~an introduction, Weather, 36;" 357 - 366. Truscott,'B.S. (1990)VSOP-NA Second Interim Report Interim Report, Augustl990, Meteorological Office, Bracknell UK. WMO .(1910) The Beaufort scale ofwind force (technical and operational'aspects) .. ,Reports on Marine Science Affairs" No.3. 22pp. WMO:(1981} final Report~ Implementation Co-ordination meeting for the Ocean Observing System D'evelopment Programme Pilot Study on a high-quality Voluntary Observing Ships' subset, De Bilt1 Netherlands, 21-23 September 1981 (unpublished report). WMO(1988) Management Committee ofthe vas Special Observing Project-North Atlantic; Report on the Second, Session. Geneva, 'S--1'December 1988. WMO~ (1990) International bist of Selected, Supp1ementary~- and Auxiliary Ships. (1990 Edition Magnetic,tape version) "WMO-41 J World MeteorologicaI Organisation, .Geneva. Wilkerson, J. (1986) Accuracy estimates of wind and wave.observations'from ships ofopportunity in the WMO'Voluntary Observing,Ship Program, in Froc. Workshop on ERS-1,WindandWave Calibration, Schliersee. FRG, 2-6 June 1986 (ESA SP-262) 'pp. 71 - 84. ,'I' ~ t -,39 - t Figure 2.1. Areas which were excluded from the analysis. Observations from the cross-hatched t areas were excluded from the Comparison Dataset. In addition. observations in the I dotted area were excluded from the analysis at the James Rennell Centre (IRC). -80 -70 -60 -50 -40 -30 -20 -10 0 i' 60 60 l I, 55 55 l l 50 50 45 45 r 40 40 r [ 35 35 30 30 l -80 -70 -60 -50 -40 -30 -20 -10 0 Figure 2.2 Distribution of observations for the Comparison Dataset as analysed at the ]RC. I The scale is in number of reports per one degree square. -70 -60 -50 -40 -30 -20 -10 o [ r-'-"__...... £...t.~L.o..L...... ~~..J....L.J..J...L....L..L~~I-J-..L.J...L..J..+.60 55 50 50 r f r 45 45 r 40 r 40 f 35 35 r r 30 ." -80 -70 -60 -50 -40 -30 -20 -10 I r ( I -40/- 1·0000 9000 8000 7000 6000 5000 4000 3000 2000 1000 o -10 - -5 -5 - 0 0-5 5 - 10 10 - 15 15 - 20 20 - 25 25 - 30 .1 I ~ ~ o .~ 9 ~ 9 ~ 9 ~ v5~g~ '{.. i"'--'_....,----__....,--,-_--'-_1:l_.·.m·._·.·.·.r;m·_.·.·.·m.·_·=.·_·rn._·.·.·.b·._·D·._·.·.m·_·m.·_.·.··•.·_·g.·_.·.·.·m._·.·.·m·._··.·.mt·.,'":::::::::::::::.•....m·······m········m·······t!··llII·~..•...... m.•...... r··m-I..rn-.m2 ...L..m....•.-.~'-..ET..t. IIIIII FNCZ .. E·:·:':':,:,:,:,:,:,:,:,:,:,:,:,:,:,:,:,:,:,:,'·:·:·:·f:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:··,:,:,:,":':':':'.' FNEF •...... ·+········;·;·;·2·;··..·····················i....·;·· •••••: ! , FNFD OIl t.·.·.·.·.·.·.·.-.·.·.·.·.·.·.·.·....·.·.·.·.·.·.·.·.·. FNGM ~ o FNGS .... I·.·.·.·.t•.•.....•..•....••...••.••...... ••.•·t·.·.·.·.·.·.·.·.·.·.·.·.•...·•·.·.·.· ·.·.·•·.•.·.·t.·.·.·.·. , FNOY .. ,.:.;.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.;.:.:.:.:.:.:.:.:.; DDLN -r------m..:·.m•.•:..·.m.:·:.:...m:·:·:.:·m.·:.:.:·~:·:···:·m····:·: G····m········m········m·····.·.m·.·.·.·.·~.·.·.·.·_.....·.· ODUC .. E:·••:.:·....:·:···.·:···.·.,,·:·,,:·.·:·:·:·:·.·.·.·· :.:.$••••••••:••••••••••-}••••;.:•••:.:•••••:-••":4 DHNE ~:·:····m.... ·:··m········m········m········m········~······m········m...... m··....··m········m...... m········m······B'l'fJEmBBm:liID_ _ ..__...... - - DHRG .. m:.... ··~·······m·······~··.... im·······~·······m·······m·····E·······E··....·E···.... ·E·······~·······E·······~··....·[I··· ..__E:::··....·~··,,···~·······m········m·······m;······m·······q-··· DIMe .. DNBR .. I·:·.·:·:·:·:·:·:·:·:·:·:·:·:·:·:·~·:·:·:·:·:·:·:·:·:. DNJR peEL 1.. r------rn:·:·m:.:·:·:~.:.:.:.~:.:·:·:~.:·:.;·:m·:.:·:·m:·:·:·;E·:·:·:B···· ..m·····:6·······B:······m········m·······E······g··· B&lm.m:razmmB·m···~···~········~·······m·······~·······~· :;r.. ----t- PELT PELU PGDG .. PGDS _ PGDW· ... •..•••••••••••:•••;••{•••••••,.;.;••••••••••••••~••¥ - PGEG LIMA e6DS GBSA _ GBW .. GBVW .. GJMR GJMS. GXES .. GZMM .. VRAZ ill VSBV3 .,.~·.·.·.·.m·····.·.E········E·.······E········E·.·.·.··sm······~········~········~········~········E·....···E········9····------IBPAT.. ------,!!·..... ·m.·",..~·.·.·.·:m·:·.·:·;m·.•:·.·~.·:·····m········E····~···.... ·m·······m········~····;···m·······s········m·····a:·······E·······m·······:m·.... ··~··:·····m·······m·.... ·m·····;·m········m:·:·:·:;m·:·:···:m·:·:·:·.rn··:···:·~:·······t: .KRLZ _ KRGJ. KRJL • KRJP KREB WMLG _ WEFZ•• wPHZ • ~ .;~ ~ WPVF •••: •••••••••••••••••fl ••••.; ••••••••;;•••••••••••••••• ·•••••••••• •••••••:: ••••••••'.io•••• WPWH • WS'£)G e+- ..,... -.... -.- ~ , __ '.' .. ~. ~c ~ I I.L ···6 cb· c$. • J cs ModelSeaSurface-Temperature{?C) Figure'B·.1a{top}- HiStogram •• snowmg".ut1IIlber:>of··VSOE observatioI1S;formodelsea sunacetemperature ranging.from -lOto3'O···OC FigureB:lb.(Oottom) ~ Numbers.ofVSOP observations byshipfOf' modelseasunacetemperature.ranging from IOto30°C: Shading indicates numbers of observations·per 5°Crange. - 41- VSBG8 t DIDA FNCZ Ff\EF t FNFD I FT'.JGrv1 Ft\GS i: Ff\OY DDLN I. IXXJC []-NE [l-iFG l DIMC DNBR l DNJR PCEL PELT PELU FGOO 0 PGDS PGDW RE] l LIMA C6DS I GBSA GBW r GBWJ GJMR - • • GJMS 1-0-1 GXES l GZMM VRAZ e VSBV3 e IBPA KRLZ ( KRGJ KRJL r KRJP l KRPB WMLG WPFZ J-D-I WPHZ l WPVF ~ ~ 0 WSfX3 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 2.5 3.0 Mean Sea Surface Temperature Difference from Model (OC) r e Bucket r ( o Enginelntake • Hull Sensor ii Unreported f Figure 6.2 - Mean seasurface temperature difference from r model (OC) by method· (bucket, engine intake,hull r sensor or unreported) by ship callsign. If more than one method was used on a ship the values are r shown separately r r - 42- 10000 9000 8000 7000 6000 5000 4000 3000 2000 1000 o LOLOLl)LOLOIt)Ll')Ll')lOLt)lOLl)LC)lOLOLOU1lOLOlOlOLl)lOLl)LOLOLOLOlOlO ~ et) C\i ,...: 0 mcO r' <0 ui ~M C\i ~ 0'0 ,... c.i M ~ Lri <0 r...:cx:i ai 0 ,..: ~. M .~ ,...,...,...,...,...1 1 1 1 I 1 I I 1 1 ,...,...,...~ I 1 1 11 1 1 1 ~ ~ ~ b bI b t~~~~~~~~~~=====~~~~~~~~~~~~::r VSBG8DIDA '= FNCZ -:,:-:.0.:,:-.:-:,'.:.:-:.%.:.:.:..:.:.:1 FNEF ,..•...... , I.·.·.·r.·.·.·.'.·.·*·.·.·.- 0.' ,, tit :_;.;.;_; em····j···;··j············j······ol ,, J FNFD t.·.·.·.·.·.·,·.·.·.·.·.·.,·.·.·.·.·.·.·I.·.·.·.....•., , $••••• ····#······t············j·············,·······...•.$ FNGM 1...... •...., 1 1 ·.·.·.,·.·.·.·.·.·.·1.·.·.·.·.·.·.,·.·.·.·.·.·.t.·.·.·.·.·.·,.·.·.·.·.·.·.1 FNGS E.·.·.·.·.·.·j.·.·.·.·.·.·.I· ..·.·.·.·.·.·,.·.·.·.·.·.·.,·.·.·.·.·4·~J FDND.'LNOY ·,-----mm--mm!!zm~ml.:.:~.:.:...~:.I:·:·=:···:·:m·:'..E:':'.:':':':.'¥'.'§.-_**.M·:·:·:···:·:+:·:·:m·:·:·:·j~~mJ~------r . ... )1 to:.:.:.:.:.: .:.:.:.:.:.:. :.:.:.:.:.:. to:.:.:.:.:.: .:.:.:.:.:.:. :':':':'0':': ••• a.... m.•••••••••.mrn:.:.:.:.:.:. :..•....:.:.: ...•...•.••...'.....'.....••...... ODUC ,:.:.:.:.:.:.:t.:.:.:.:.:.:f·.:.·.·.~;.#@ii.· .••;.;•••·.I:.:.:.:.:.:.:t.:.:.:.:.:.:J ~ DHNE ~ t:.·.·.·.·.·.·,.·.·.·.·.·.·.,·.·.·.·.·.·.·I.·.· ; ..%.$ &*** ,·.·.·.·.·.·.·t.·.·.·.·.·~·,·.·.·.·.·.·., ,············t············J··········..··1 DHRG ~ F········~···I·.·.·.·.·.·.·E.·.·.·.·.·.·¥·.£$# ••.. ·.·.£·.·.·.1············1············,········....., mt····~ m·····I1§f·~······~·····EJ···~······m···1~· ·~·····~+·~ DIMe E.·.·.·.·.·.·I.·.·.·.·.·.·., ....·illl···¥iIi······m..···::I·__e····...... ·m·....·1 DNBR Imm I:·:·:·:·:·:·:'·:·.·:·.·:·.·f:·:·:·:·:·:·'· •••••••• 1.·~.·.·....f:·:·.·:·:·:·*·:·:·:·:·:·:J·:·:·:·:·.·:·1:·:·:·:·:·:·:1 ~NCEJRL l------~mml::·:·:i·:·:··;-.~t~:.:.:~.:.:.~;.:,.::.:. :;.:.:.::+:..;..:.:.;:.;.1;:.:..:...:.::.:,.;....:.:..:.:;.f:.::.....;:.;.*::..;..:.. :::.:*::':':';:':'::'1-::.:.:.:;.:.:1:.:.:.::.:.:;-:-1:;.:.:.::.:.:.:*.::.:.:.:;...:*::.:.;::;.;.1::':':'i:':':'~:' ~~::.:.::...:.~_+ . .., .:.:.:.:.:.::.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.'.:..:.:.:.:.:.:.:.:.:.:.:.:. :.:.:.:.:.;..:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.: :.:.. : : : PELT ,.·.·.·.·.·.·.,·.·.·.·.·.·.·I,k····;·····f*Wi.·.·.;.·.·.·J.·.·.·.·.·.·.,·.·.·.·.·.·.·, PELU ~ ~ 1·············1.·.·.·.·.·.·., ~t····~·····;:J···I·~·....·m·····]J·.mlB]·~······~·····t~···m·....·m:··lm·····m....··m·r···m·····m·····! PGDG I.·.·.·.·.·.·.,·.·.·.·.·.·.· ...... £ j ~ PGDS ,············f············ ·.·.·.·.·.·.f.·.·.·.·.·.·1 Z PGDW f:·:·:·:·:·:· :·:·:·.·:·:·:t·:·:·:·:·:·:J PGEG '·:·:·:·:·:·:·1:·:.:·:·:·:·:(·:·',:,:,:,:, ":"':':"':1':':':':':':+:':':':':.:" LIMA C6DS F···.···.···.t····.·.···.·I······.·····.,·······.·.·.·t.·.·.·.·.·.· •.i§{ri.·.·.·.·.·.·¥·.·.·.·.·.·.J·.·.·.·.·.·.·I.···········4 GBSA E············Pies ·.·.·.·.·.·.·I.·.·.·.·.·.·J-.·.·.·.·.·.I·.·.·.·.·.·.·1.·.·.·.·.·.·:1 GBW ,.•....•...•.£...... •...j ml·····m....·~···t.~·.....~·.·.·)§. llIlllII ·~·,···m······~t-··m·····m····J ~ GBVW I:·:·:·:·:·:·: ·:·:·:·:·:·:·f:·:·:·:·:·:·*·:·:·:·:·:·:1 GJMR 1.·:·:·:·:·:-:[-:':':':':':+:':"':':':" •••:.:.:.:.:.*.:.:.'.:.:.: :.:.:.:.:.:.:t..:.:~.:.:.:J.:.:.:.:.:.:.,:.:.:.:.:.:.:[.:.:.:.:.'.:+:.:.:.:·:.:·':·:.:·:.:·:·:1·:.:.:.:.:.:., GJMS m':':'~:':':~':'*~"':"m"':'3: :m·:·:·:~·:·:.tE·:·~:·:·:·m:··j ~ GXES ~ F············j-············I.·.·.·.·.·.·.,·.·.·...... ; ···.·.·.·%.1·········:··1·············,·······..·.·.·1.·.·.·.·.·.·,·.·.·.·.·.·.1 GZMM I.·..·.·.·.·.·.,·.·.·.·.·.·.·t.·.·.·.·.·.·;·.·.·.·.·.·.¥· ..·.·.·.·.·.·I.·.·.·.·.·.·.,·.·.·.·.·.·.· . Wi·.·.· ..·.···.. I.·.·.·.·.···I.·.·.·.·.·.·.,·.·.· £....•...••..£ J VRAZ mi····m······m···'~······mi····~.f.im§·i.E#;.. O···m······m4m.... ·~·····m,.J.·m······:!m·····1 m!m VSBV3 ~ E.·.·.·.·.·.·1.·.·.·.·.·.·.I·.·.·.·.·.·.·•.·.•.·.·.·..·*·.·.·.·.·.·.1·.·.·.·.·.·.·1.·.·.·.·.·.·.F·••••.••••••· M·.·.·.·.·.·.·,.·.·.·.·.·.·.,·.·.·.·.·.·.·•.·.· * J· .•.·.•.·.•.·I ~ IBPA :.: : :.: :.:.: :.: :.:.:.:.:.:.:.:.:.:.:.:.:.:•....•..... :...•...... KRLZ ~ KRGJ ~ I·.·.·.·.·.·.·,.;.·.·?·.·. ·.·.·.·.·.·.·,.·.·.·.·.·.·.1 ~ ~ .:KRJL ,·.·.·.·.·.·.·1···········:1····..·······, ,········,;·····F············*··········.-· .·.·.·.·.·.;.1:.·.·.:.·.·.·1.·.·.·.·.:.'£·.·.·.·.·.·.". ·.·.·.·.·.:I······.·····l·:·········j'~ ~ KRJP mm !.·.....·.·.·.,·.·.....·.·.·I.·.·.·.·.·.·!·.·.·.·.·.·...... p.J,. ¥....•...... ,...... •..·.·I.·.·.·.·.·.·.(·.·.·.·.·.·.£.·.·.·.·.·.·j KRPB ~ ,·.·.·.·.·.·.£,·.·.·.·.·.·1.·.·.·.·.·.·.,·.·.·.·.i.·.,f.i...•...... ; . §.; +.·.·$.·.·.·;..·.·.·.·.·.1·.·.·.·.·.·.·1.·.·.•.....•.,.....••.1.t.3 WMLG WPE'Z m;m F:·:·:·:·:·:·J:,:,:,:,··:·:J.:.:·:·:·:·:·E:.:.:.:••·:·t·:·:·:·:·:·: ;.:{§#.:.:.:.:.:.~j.:.:.:.:.:.:.I:.:.:.:.:.:.:t.:.:':':':"+:.:.:·:.:·:·$:·:·:·:·:·:·:1·:·:·:·:·:·:·1 ~ Wl?HZ ~ mm! rnt.:.:~.:.:.:;];.:f[l:·:·:·m:·:·:E·J:·:m·:··,~:,·.:il!i·},·:s3:.aB"·BI IIIIE:·:·m··:··!i·:·*m·:·;·l'i··:·j·m:J-:m·:·:·:m·:·:·BI:·:·m:·:·:;[;1···:l~·:·:·m:·:·:·:3 m;] WPVF E.·~·.·,·.·.··£·.·.·.·.·.·.j·.·.'.·.·.·.·I.·.·.·.·.·.·.F·········.·.t············E·~· ..*....·· @§·.·.·.·.·.·.··j.·.·.·.·.·.·.,·.·.·.·.·.·...t.·.·.·.·.·"J ~ WPWH ,·.·.·.·.·.·.·I·············F············¥·····...•..• 'WSES I .~ b '0 VSOP ·_·.··ModelSea ··SurIac;e'Te"I1lperatl1fe,(O,C) Histogram.··••showmg;nllIl'ftler···.of·VSOP 'observations··.differmg·from.·.model·sea:sUIface ';1(}flf temperatures~ging.··from-15··.to ····I5°C 'SOf ·6·" ·FiQllke···6:3b~'{bottorij) .. - Hlllmber:s··ofV$(IJP.···.observationsdLll'ering·from 40 .mooel"seasu1"face::terqperatures··byship for te~pe.ratur:etiJffereno:esranging:·from·-15.tolS·;9C. 2<0"tii~~~~~;i;~; "8haamgiindicates··~;ers;Qf,observations.per.1~G ;Tange~ t l - 43- l ...-..- 30 0 0 l ~ Bucket Q) 25 ...... _...... I .. Engine Intake ..::I ..as ....-._..- Hull Sensor Q) 20 t a. E I, Q) l- 15 Q) I (.) ( ...... as 10 en::s 5 l as enCD O D- 0en > -5 - 5 0 5 1 0 1 5 20 25 30 r Model Sea Surface Temperature (OC) ( ,.....". 0 0 ~ 4.0 ..CD l ::s • Bucket ..as 3.5 .. ·········0······· Engine Intake Cl) c. 3.0 f E ...... Hull Sensor Cl) 2.5 I- [ CD (.) 2.0 ....as .. 1.5 en::s ~, as 1.0 Cl) ...... •...... , .•_ ....rr"...... ~ en 0.5 .1::2•••••••••••••• •..•• G) 0 "C 0.0 0 :E -0.5 D- -1.0 O en - 5 0 5 1 0 1 5 20 25 30 > r Model Sea Surface Temperature (OC) t r Figure 6.4 a (top) VSOP measured sea surface temperature binned on model sea surface temperature separately for temperatures measured by bucket, engine intake and hull sensors f Figure 6.4 b (bottom) Differences in VSOP measured sea surface from model values r binned on model sea surface temperature separately.for temperatures measured by bucket, engine intake and hull r sensors r r f" -"44- a20 ...-...... -.....--..~-.---r--...... ,.- ...... -..-....,..-....-...... - ..... •• ··Bucket 2';5. ········~O·······Enginell1take 2.0"; _...... _. HuU··S.ensQr '1.5" 1J); '0..5,. o:·o·~ ..0.5·..••· -1.:0· ~t~5, a.. ..2LO . -.l._....oIo-_..&----.ilr....--.l._...... _.....-.__..r....-...... _.....-._...... -_ '0 -10;0 :0 c>UJ ;.)F~~e ·;:6J5·~a':r'tqPJ··- VSOPmeasured·.:sess.t'lffaceiterr.1peEaturebiJ0ued..mnIsemmle··T '··latent,·'neat·'ftm·.{;calcma:ted'.!f1'\G};mbrifk:'·!ormmae'}_g"VSOP .• p:ar;ar.Fleters·.;e;xe~ptIar:.·m0del,·:sea·:surl'ace:temperatt.lX~} s~paratelyfClr;teroPeT,aru.r-es.measmedbY'·bncket· ··engiI1eintake anti~b:tlll;senseFs D!illerl~'1'iCes·:mb;tl0l -0.4 ..... 53 7 o 200 400 600 800 1 000 rI l Total Cloud Cover (oktas) I Short Wave Radiation (W/m2) r Figure 6.6 - VSOP measured sea su.rface temperature binned on short V'/ave radiation for daytime measurements, and total cloud cover for nighttime measurements, separately for temperatures measured l by bucket, engine intake and hull sensors 3 r ...... 2 0 ...... ,0 UJ... U) Engine Intake t Q) 0 o Engine tntake 0 o ":E 0"" • Hull Sensor -1 D. 0 Hull Sensor U) > -2 I -3 r 0 123 4 567 8 r Depth of SST Measurement (m) Figure 6.7- Mean differences in VSOPmeasured sea surface temperature f from model values plotted against mean sensor depth for r engine intake and hull sensor measured data r r r r 3.0 --0--- <200 m 2.5 _._- _- 6 .. >200m .....,0 2.0 I- Engine Intake U) en 1.5 -; ..".."..".;,-.." \\ '0 1.0 0 \. ,.-i :lE \. .,' 0.5 ······~~···:O·····_····o ..".."•., c.. '-, ~:-_- •.."...... 0 ...._..'0 0en 0.0 > -0.5 -1.0 -5 0 5 1 0 15 20 25 30 3.0 •••••~... <200 m 2.5 6 --..-- >200 m ...... 0 2.0 .... Hull Sensor tn en 1.5 G) "C 1.0 0 ~ c.. 0 (IJ > -0.5 -1.0 -5 o 5 10 15 20 25 30 Model Sea Surface Temperature (OC) Figures 6.8 a. b.·c - Sea surface temperature difference from model (OC) by method (bucket. figure 6.8a top, engine intake, figure 6.Bb centre and hull sensor, figure 6.Bc bottom) for ships greater than and less than 200 m length l ~ l - 47 t 10000 9000 t 8000 I 7000 6000 t 5000 4000 l 3000 l 2000 1000 lI o .. 10 ....5 -5 - 0 0-5 5 - 10 10 - 15 15 - 20 20 - 25 25 - 30 ..!.. I ~ ~ ~ ~ o ~.. ~ ~..? .~ ? .~ VSBG8 r:.:.:.:.:.:.~ m.._ m.._...... •._ ·_.··.·E._.·.·•.·_.·.·.·.m·_.·.·.·.m·._·.·.·.m·._·.·...•.._...... •.., ::::::::::::::::.::::.:.3..•...... m..•...... mm..•...... m·II·······g··.··· DIDA .. ...•._=.._ .:.:.: :.:.:.:.: :.:.:.:...;.:.: :.:.:. ;;;;;;r;. FNCZ .. E·····?·:·;··...... ·:···:·:·········;·:·:·:···:···3 I:···:·:···:·:·········:·:···:·:·;·:···:·:·:·:··.··:·;· FNEF , +=.:••'§ 11< at &. ?.,t!,¥*;...... •...... j FNFD _ , ,.**.# ?..;.,.*~s.. f!; ••••••••• . f FNGM FNGS .. , J ·.·1.•••.•.••·.•.·.•.•.••••·.·.·•••••·.·.·.$·.·.·.·.·.·.1 FNOY" DDLN .------....·...·.m·.·.·.·m·.·:..m.·.·.··m·.·.·.·m.·.·.··m1.. · m··..m········m········~·······m·······m·.·.·.·.m·.·.·.·.t-· r ODUC .. :.:.·.;0·...:.·.·.·.:,·.:.·.·.·.·•••·.:.:.:...·.:.:.·1 • g~ ~ '.. li!·····.m··m······.E·······m······.·m··.····E··:;m::m:::::::E::::::&.::::.m::::::::m:..::::m:::::::E::::::::11:11 1111 DIMC .. B······m;·······m·····m······g····..·m····m·······D····-.1\------E'..DlIiD .a·····m·······m·······g·······=·······m;·.' '.' ...•....;mq-. DNBR .. mF:····m·····$;oS!····;;B··;.···g>.·····i!l:·:·:···m·······0····ElBJI&aI.. E··m·······:B·;;····m··:····m···$··B:·.•$.m;.:•••:.g..:o••••, ~ DNJR .. ;.: :.:.: : :.:.:.: :.:.:.: :: ;':':':':"':':':.:.:...:...:..':':':':':':':';':':': PCEL .. :.: :: :.:.: :.: : :.:.:.:.: :.: :.: :.: :::...... PELT .. m······m;:·······m········m·....··g·······s·....··m········;:m·····s·······g·······g········m·······g·······g········m·······lJI· B:··m;········m·······g·······s·······g·······s·······m;....···;r PELU mD!..··... ·I1... ·mum...... m..... m··.. BBl······E······m.. ···i!·...... ···g··..B···....m···D· ·m········m·······m....·••··•·B······..· PGOG m ··!!········;:m·······m········m········m·······;:m······..~. ~····m·······m m·······m········m·······m . PGDS _ m······m·······~········:m·······m·······m·······m·····•••g... IE···m···..··m··D····..~..···g·m:··m·····r PGDW Emf:·:···m··:·:·::,m·····:··m·;.:·:·m!····$:·m:···:··m"':';.91·'; ··&·'2·&2·miRE·fang·s-mM ~ PGEG ..:..:.....:.:.:...:.:.:...:.:.:.:.:.:.:.:.:.:.:.:.: .:.....:.:.:.:.:.:.:.:.;.:.:.:.:.:.:.:.:.:.:.:.:.:.;.. r LIMA r C6DS • , •••••••••••••••••••••"•••••_•••••••••••••0•••••0•••••t ,', GBSA _ GBW. GBVW • GJMR .. GJMS GXES GZMM • , ...... , , i - VRAZ .. I '··m········m········E········E·········m·.······m··..····m···m·~~m. VSBV3 _ IBPA .. KRLZ KRGJ • KRJL .. KRJP .. KRPB • Cl)« WMLG _ ::> WPFZ' Wl?HZ • Wl?VF • r Wl?WH WSDG ~ fII-IP ~ -PI f _M§•• .§ ,_ ~ ~..·.§ • (.§_•• s'_"$'..' -... ~ r ~ ~ ~ ~ ~ ~ I ~ 0 ~ 0 ~ r Model Air Temperature (OC) r Figure 7.1 a (top)- Histogram showing number ofVSOP observations for model air temperatures ranging from -10 to 30°C f Figure 1.1b (bottom) - Numbers ofVSOP observations by ship for model sea surface temperature ranging from r -10 to 30°C. Shading indicates numbers of observations per BOC range. f r' ! - 48- 6POO 5'000 4000 3000 2000 1000 -15- -10-10- -5 -5-0 0-5 5 -10 10 - 15 15 -20 20 -25 25 - 30 I u o VSBG8 : :.: :: 'r :.« DIDA • .':':'.':':':'.':':':':':':':'.':':':':':'.':': ;0••••••••••••••••••••••••••••: :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:. FNCZ ,.:.: :.:.:.:.:.:.:.:.:.:.:.:.: :.:.:.:.:.:.#:.:.:"':':':':':';':':"':':':':':':':':':':':!:':':"':':'••:.:.:.:.:•••:.:.::;.:.:.:.:.:.••:1••••$...:.;.:.;••••0:$.g.;?••:.:••.;.:.:. .:.:...:.:.:.:.:.:.:.:.;.:.:.:.:.:.:.:.:.:.:.:., FNEF _ ·i •••••••••••••••••••••••••••••••••••••• ••••••••••• . FNFD . FNG! t··············································,·······.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·1.·.·.·.·.·.·...... •...... ,...... •...... •....·.·.·.·.·.·.·.·.·.·.·.·.·.·.·1.·.·.·.·.·.·.·.·.·.·.·.·...... •., FNGS .. t·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·E.·.·.·.·.·.·.·.·:·.·.·.·.·.·.·.~.·.·.·.· ....·.·.I·.·....·.·.·.·,·.·.·.·.·.·.·.·.'.·.·.·.·.·.·.·.·1.·.·.·.·.·.·,'.·,·.·.·.'...·.•. ·.·.·.·.·.·.·.·.t~ FNOY· DDLN r---m·:·:·m:·:·:·:·:m;....·:·:·m:·:·.···m:·:·:·:·.m···.·:·m:·:·:···m:·:·:·:·:m;·:·:·:·:·m:·:·:·:·;m:·:·:·:·:g·:·: .. m.·.·.·.m... ·.. m·....··m.·.. ·...E.·.·.·.·g.·.. ---+ DDUC ~ D~ :: . ...:...••...... ,.,., ', ,.,..'.•. DHRG . DIMC .. DNBR ,.:.:.:.:.:.:.:...:.:.:.:.;.:.:.:.:.:.:.:.:.:.:. ...ff·.·.·.·.~·.·$.·.·.·.·.·.·.~·.·.~·.·.~, ~NCEJRL J_~--~':i·:·:·:·:··:·:·:·::·:·:·:i·:·~:·::·:·:·::·:·:.:.::.:1 :..;.;..:.;:;#:..m:.§..: -J:. s; J :.:.:.:.:.:.:.:.:.:.: :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:======:-:., :.:.:.: : 0:••••:• :.:.;.:.:.:.:.:.:.:.:•••••:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:. __ ..., ,'...... •.'...... •...... ',...... •...•....'.....•.: , :,., , , ' , '..,'.'..,' ,•... :.',..', :...... •'.., ',,',: PELT ... PELU ... I?GDG ... , ! 1$••••••••$••••••.•.••.••'••.•••••••.•••,•.•••. .·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.-;.·.·.·.·.·.·.1 PGDS _ ,', * '.' ,., , . . g , j PGDW .. 1:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:·:1·:.:.:.:.:.:.:.:.:.:•••:.:.:.:.:.:.:.:.:.:.:.:.¥••••* . ...g , ; !:.:':':":.:.:.:.:.:.:.:.:.:.:.:•••:.:.: :-..:J PGEG ,.:..:.:.:.:.:.:.:.:.:.:.:•.• :.:.:.:.:.:.:.:.:.:.#:.":':'~:':""':':"':""':':'~:':""':. Llb1A, :: ;0,0 ' . C6DS , ,. .•••••••••••:•••••.•••••.•••••:J GBSA _ EF·Mm·······g·····..g····..··m·;:···g·····#m·····__IIlllIIIllllII IllIlIIII_m······g····#·g········g·······§$····.g•.•..£§.....j GBW.. I....·.·.·.·.·.·.·.·.·.•.·.·.·.·.·.·.·.·.·.·.·.· GBVW .. ,.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.,.;::..>.•••••••••••;.;•••••••••;>,••••••{••••• GJMR .. c·:.g··m.;...:...@•••••.;.g.•.;•..•. G~ 1:·:·:·:·:·:·:·:·;·:·:·:·;·:·:·:·.·:·:·:·:·:·:·:1·:.:.:.:.:.:.:.:.:.:.:.:.:.;.:.:.:.:.:.:.:.:.:. ~ GXESGZM4 ' . VRAZ ml···:·~···:m·······~·······:m···~·······:m···ml·.···m··.... ·~·······:m·······~·······:m·······~·....··•·mB§B.mmIJ e····:;m:·······m·······:;m:..·····m;·······:;m:·······~•••••...J VSBV3 . IBPA ... ':'.':':..':'.':':.:':.:'..:.:'.':':.:':'.':':':':':':.:.:.:.:.:'....:.:.:•....:..'..:...:...:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:':.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.: KRLZ. KRGJ ... ,..:...... •...... '.., ...§.••••....•...•.•~;.....•...:.....•..•..•.. .~ j .. KRJL ...... ' ~...... ,•...... "'....•...... •...•...." '.' s·.·.·.·.·.·.·.·.·. ',' , 'iA. ••: ~•• KRJP I Figure7.2a· (top) - Histogram showmgnumber ofVSOP dbservationsfor rnodeldew point temperature ranging from -15to,30°C Figure 7.2b (bottom) - NumbersofVSOP ol:>servationsby ship for modeLdew pointtemperature'ranging from -15 to 30°C. ,Shading mdicates numbers of observations per 5°C range. l l - 49- f -80 -70 -60 -50 -40 -30 -20 -10 l 60 I t 55 55 t 50 l r. 4-5 4-0 40 2.0~ d:g~ I -1.0// t 35 35 -2.0, I I l r .30 ~ -80 -70 -60 -50 -40 -30 -20 -ID o -80 -70 -60 -50 -40 -30 -20 -10 r 60 f 55 55 I 50 4-5 4-0 40 2.0~~ d:g~ -1.0// 35 35 -2.0 I I I 30~~~~~~~f-+olI~r't'rC+-f-'tI"l~f"&r&~~"""'''''''' .30 -80 -70 -60 -50 -40 -30 -20 -10 o Figure 7.3a (top) - Air temperature difference from model (OC) averaged in 5° squares over the North Atlantic. Figure 7.3b (bottom) - Dew point temperature'difference from model (OC) averaged in 5° squares over the North Atlantic. -50 -40 -30 -20 -10 ci:g~~·..•.· .••• :.··•.••.• :~~ -o.s·#).0.· 0:...... ••. ....•. •...•...... /Z.. :...... •...... •...... •.•• -to, "I -70 -60 -50 -30 -20 -10 o -sn -70 -60 -50 -40 -30 -20 -10 60 55 50 45 1.0~ 0.5·o.·.·~.· -O:5~o. ..···.··.·i -1.0, I , -70 -60 -50 Figure 7.4a (top) - Summer (May to September inclusive) .air temperature difference from model'(OC) averaged in 5°' squares over the North Atlantic. Figure 7.4b (bottom) - Winter (November to March inclusive) air temperature difference from model (OC) .averaged in 5° squares over the North Atlantic. 50 4-5 45 4-0 40 r g:g~1.0~ -0.5 / ~ -to. ~ ; 35 35 f r 30 -80 -70 -60 -50 -40 -30 -20 -10 o Figure 7. 4c (top) - Nighttirne Summer (May to September inclusive) air f temperature difference from model COC) averaged in 5° r squares over the North Atlantic. r Figure 7.4d (bottom) - Nighttime Winter (November to March inclusive) air r temperature difference as in 7.4c. r jJ I -2 ~ ~ ~ .... - 1 to 25 .....------.. o '\lSOPAir .u...•....······ModeJ.Air --0-- \lSOP 'Dew SPaint .,...... " ~o • :Moool·'[);e,wPoint o .,-"" 15 1'0 :5 -+-...-...... -..---..--....-.....-...-...... -~...-...... r---I--....-...-...-...... -...... -...... 0 3'C):6'O'9;Q 1;2'01Sl,).18:'021!024'02'1~03{O:03303;6:0 Figure 7.4e (top) - Differel1cehetweellVSOP and model airtemp'er:attmes by-season (Spring, March. toMay';Summerl Jooe to August; AutumnI September·toNovember;an:d Winter,Dece~,ertofebruary}. Figure 7.4f (bottom) ~ VSO,E and·model air and dew point.temperatures averaged in 30 day sections. ;~ i , \.,--_.~....---',,--_.;;.;-----..-.-_...... ~--_._"-:'V--_ ~':~_ ---..~~--,,·--~ .•t '" \---;..'~---r---w , .... "" '" 'fI -.s~-' --""..~-"'~r--" .. .. ---. VSBG8 V1,~~ IIII '-T' 1- - III DIDA v~gll I( 0/7' .... 1- 1 I I I FNCZ FNCZ FNCZ FN3= Ff\EF RH' FN=D FNFD FNFD FNGM FNGM FNGM FN3S FN3S Ft\GS Ff\OY ~ ~:xI. ~ J III -'-::1"-1 I III DDlN I• I I ] ] 1 I I COJC OOJC COJC CH\E ll-l\E D-N: [)-ffi [)-fR3 [J-fG DIMC DIMC DIMC DNBR DNBR DNBR DNJR ~~ PC8 ~ I 'F"~_ A _ I I I III ] 1 'IX' II - Ix - II PELT PELT PELT PEW PEW PEL.U FG:X3 R3CG R?OO PGDS PGDS PGDS PGfJN PGDW PGfJN RE ~I III 1rv::=1 III LIMA rvp I I II III rlmJ C6DS CaDS CGDS GBSA GBSA GBSA GBW GSW GBW GBWl GBWJ GBWI GJMR GJMR GJMR GJMS GJMS 1 I I ... GXES GXES ~~ ~ ~ GZMM GZMM I L r... I 11 I VRAZ VRAZ ~ VRAZ VSBV3 VSBV3 J--4- ~-~--+:-z.=--~-_+_-_+-__f. V~~~~ I ISPA ISPA I J. i I r -- . II t--t KRLZ KRLZ KRLZ KRGJ KRGJ 1 I...... KRGJ KRJL KRJL KRJL KRJP KRJP KRJP KRPB KRPB KRPB WMLG WMLG WM.G 'N'PFZ WPFZ 'NPFZ WPHZ WPHZ WPHZ "NPVF WPVF WPVF W?NH W?NH W?NH WSlX3...... ~,1.....1.....1 ...... , ,..-,_,-+-...... 1 '....' ....1 -F.~I,....., ....., +...... 1 II""'i"..,.'...... ""'1I~''''''I-+-...... 1 II""'i'I"'I+,.....1 II""'i',-t-·, WSlX3. I, , i , I i • , , I',,,I', ,', I ,i, , r, i , i I' i i , I',,,I WSlX3. I" I, I',,, r' la I ri, ,, I' I,I I',,,I',,,I' u , , I -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 -1.0 -0.5 0.0 0.5 1.0 1.5 2..0 2.5 3.0 -2.0 -1.5 -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 Mean Air Temperature Difference from Model (OC) Mean Air Temperature Difference from Model (OC) Mean Dew Point Temperature Difference 'from Model (OC) • Screen o .Psychrometer 6. Unknown Figure 7.5a - Mean air temperature difference from model Figure 7.5b - Mean dew point temperature difference from Figure 7.Sc - Mean air temperature difIerence frc::>m model (OC) by method (screen, psychrometer or model (OC) bymethod (screen, (OC) by method (screen, psychroIIleter or unreported) by ship callsign. psychrometer or umeported) by ship unreported) by ship callsign for r€!gion -20 callsign. to -10 OW by 45 to 50 0N for the pE~riod February 1989 to February 1990 - ...54- 8000 100(} :.6f>OD····' 5000 4'00,0' aooo, 200.0 toao .0 1O.lt)ll) Lt')1l) 1.0 1.0 LOlOIl)LOlO lOl.O 1.0 1.0 ;1l)ll)LO"lOlOLOlOl.O· LOI.OLO lOl.OLO ott:ieti>N ",:oai.cx:i ~u:iari .~roN"':.ci <0''''': cJci~t.riu:i~Cx:i m VSBG8m~1'f~::~:::::!E!~::~~======~::~::::::~:::::t~~~ ~~~~ FNCZ FNEF ,·.·.·.·.·.·.·r.·.·.·.·.·.·f·.·.·.·.·.·.,·.·.·.·.·.·.·,.·.·.·.·.·.·.';.2.;.·.;.· ············I············j·············,········....., FNFD t.·.·.·.·.·.·,.·.·.I·····~·F····t ••..oi!.. ·.·.·.·.·.·.I.·.·.·.·.·.·j FNGM ~ ·t.·.·.·.·.·.·j.·.·.·.·.·.·.,·.·.·.·.·.;.·,.·.·.·.·.·.··,···.·.···.·.I.·.·····.·.··j.···.···..•..,, FNGS ,·.·.-·.·.·.·.·,.·.·.·.·.·.·.,·.·.·.·.·.·.1.·.: j f l--....;....-,---...,md~=··:·:·:·;:··· =d'i·:·:·:·::·:·:: :.;..• ::+.. :.:;.:.:.:.:t:.:.::.:.:.:.~., FNOYDDW" .:.:.:.:.:.: .:.:.....:... :.:.:.'.:.:.: ==::.:.:.;:'.' :':':':':'.' :':'.':':':': ...:m"'----....:.-----t....:.. DDUC DaNE ,·.·.·.·.·.·.·1.·.·.·.·.·.·.1 mt·····m······m·j···~······:m····im!·*m······.. IG···m·.j§·m:·a:rn······m······E;j····m·······i:iI··,··m······~·····F~····;,~·······:1 DHRG ,············t·······'!····I.·.·.·.·.·.·.is..§.x.- .·.•.·....·.·,·.·....·..•....1 DIMe .~ EE:J ,·.·.·.·.·.·.1············10············,..•··•••...;...... , •••••••••••• , •••••••••••••j DNBR F:·:·:·:·.·:·I:·:·:·:·:·:·:I·:·:·:·:·.·:· 1 DpN..·EJRL l------!,·:·:·!:·:···:;*:·:·:····:···:J:····:··"':'=:I'~;';:"":;:' ::.1.:..:.:.:.:.::...•:.:.:.:.:~., ---;mT~~§':':':'~:' ... ., :_:.:.:.:.:.:.:,'.:.:.:.:.:.:.".:.:.:.'.:.:.:_:.:.,.:.:=====:.:.;•••t.: :.:•••••..:...:.:•..:.:.:.:.:.:.:.:.:.:.:.:. -----+ C PELT ,······..:·····F······:··.··I············£···;.....t·,#W·····:·:··..·E··..·········t···········.J l?ELU ,·..···········,········..···1············ .·.·.;.~·$.i············t············)········ ....., ~ PGDG I·············,············I······§···· ·•··.. ······t·······,;·~··j·············1 l?GDS ~ m'·····m······~·3···m······[)···· IWm~···m·······~··J··m······1E·····' PGDW mm ,.:.;.:.;.:.:+••••••,>,.$ .••••$••••·.,·:.:.:.:.:.:I:.:.:.:.:·:.j l?GI··~. 1------~~E:·:·:·:~· ~~':.;:.:.:.;.::.:E.:.::.:.:.::.:§:*:. ==='::';':'::':':il_mmr-lar-mmmmI1!ll~r--+ L -&'I1tW"'\.G" 1 : :.:.: :. eGDS mmJ rnt····m······m··J···~······m····'m·······m······IlIIlIll ..~~~~~ GBSA 1....·.·.·.·.·.,·.·.·.·.·.·.·1.;•.•...%.·. ·.·.·.·.·.$·1············1············, GBW E.·.·.·.·.·.·,.·.·.·.·.·.·.,·.·.·.·.·.·.·E.·.·••.••••• .·.·.·.·.·.·,·.·.·.·.·.·.1·.·.·.·.·.·.·, GBVW :·:·:·:·;·:·*·:·:·:·.·:·:1 GJMR t.·.:...:.:·:I.:.·•.•:.:.:. :,;':':';':.1':':"':':':4':':':':':':" ~ GJMS 1:,:,:,'.:.:.:1.:.:.:.:.:.:· GXES rnt····m·······m·t'·~·....·m····'m·....··m······mfi,·.:i.j.",.••g.. 1IIIl!llI llilIE···m·····i·:l:i··*·~·....·m·····j ~ GZMM I.·.·.·.·.·.·.,·.·.·.·.·.·.·E.·.·.·.·.·.·I·.·.·.·.·.·.1·.·.·.·.·.·.·,·.;.·$·.·. ·$·§····:··,············I············j··,:· ,, F············I·.·.·.·.·.·.·) VRAZ E············I·.·.·.·.·.·.·, EF·.·.~·.·.·.~·.· mi -.::;··£m······::I··f·~······~·····Jrn·····rn······m··'··m·······m····1 VSBV3 ~ ,·.·.·.·.·.·.+.·.·.·.·.·.·1.·.·.·.·.·.·.1·.·.·.·.·.·.·f.•. ·.·.~.·.·*·.·.·.;,.·.·.t·.·.·.·.·.·.·i.;.·,·.ili.'. .·.·.·.·.·.·j.·.·.·.·.·.... ,·.·.·.~.·.·.·f.·.·.·.·.·.·4 IBPA :'.':':':':': .:.:.:.:.:.: :.:.:.:..:.:..':':':':':": .:.:.:..:.:.:. :.:.:.~.:.:.: .:.:.:.:.:.: I Figure 7.6a.(top) - Histogram showing number ofVSOP observations differing from model air temperatures. ranging from -15 to 15°C Figure 7.6b (bottom) - ·Numbers.of VSOP observations difIeringfrom modelair temperatures byship for temperature differences ranging from -15 to 15 °C.Shading indicates numbers•ofobservations per 1°C range. [ - 55- 4000 3500 l 3000 I 2500 2000 t 1500 l 1000 I 500 f o ~~~~~~~"~~~~~~~~~~~~~~~~~~~~~~~ ~cnt\i"":oaieO~u:iLCi~cnN"":oc:i~NC'?~Lri Figure 7.7a (top) - Histogram showing number ofVSOP observations differing from model dew point temperatures ranging from -15 to 15 °c 100 f 80 Figure 7. 7b (bottom) - Numbers ofVSOP observations differing from 60 r model dew point temperatures by ship for 40 . ~$~ temperature differences ranging from -15 to 15 °c. 20 \~~~~I~~l~~;~;~~~ Shading indicates numbers of observations per lOC f range. r f Screen 25 Psychrometer PRT 20 Cumulus « Q. 0 o tn > -5 -10 ...... 1'.0 ,,'5 0 ;51015 20 25 Model Air Tempe:ralu're (OC) • Screen ..CD ········0··'···· Psychrometer .....::J ---"0--· PRT ..ca 0.0 _.'_ -.- CD .. Cumulus Cl. E Cl) t- ..0.5 "'.0 ,Q. o (J) ...... ,~ ~ > .. 1.5, .J..AI ...... 10 .. 5 05 1015 20 25 Model ,Air Tempe'ratu;re (OC) Figure 7.Ba (top) - VSOP measured·air temperature binned on modelair' temperature separatelyJar temperatures measured byscreen, psychrometer,PRT and bythe Ocean Weather Ship· Cumulus. Figure 7.8 b .(bottom)- Differences in VSOP measured air temperature from model values binned on model air·temperature separately for temperatures measured by screen. psychrometer, PRT and by the Ocean Weather Ship Cumulus. l l - 57 - l, ...... -. 30 0 ...... ,° Screen l. 25 CD I a- Psychrometer t :::I ...as 20 PRT a- CD l a. Cumulus E 15 CD l I- 10 ...c l ·0 t D. 5 ~ CD 0 C D. 0 -5 en > -10 -1 0 -5 0 5 10 15 20 25 30 Model Dew Point Temperature (OC) f ( I o...... , ° 6 ..-.- ...... CD l ...::s 5 Screen ...as • CD ·········0······· Psychrometer c. 4 E ---fr- PRT CD 3 I- ....•... Cumulus 2 o.....------..-..-.--...... ~~------I -1 -2 -3 D. o -4 ~ en -1 0 - 5 0 5 1 0 1 5 20 25 30 > Model Dew Point Temperature (OC) r r Figure 7.9 a (top) - VSOP measured dew point temperature binned on model dew point temperature separately for temperatures measured by screen. psychrometer. PRT and by the Ocean Weather Ship Cumulus. Figure 7.9 b (bottom) - Differences in VSOP measured dew point temperature from r model values binned on model dew point temperature separately for temperatures measured by screen. psychrometer. r PRT and by the Ocean Weather Ship Cumulus. [ [ - 58'- 3000 .------~----~~~------..,. 2500 2000 1500 1000 500 0 Lt) 0 Lt) 0 Lt) 0 Lt) 0 Lt) 0 Lt') 0 Lt) 0 Lt) 0 C\I C\I et) et) v v Lt) Lt) co co ...... co 0 1.0 0 Lt) 0 Lt) 0 LO 0 Lt) 0 L(). 0 LO 0 LO C\I C\I et) et) ..q- V LO LO co co ...... 3000 ..------...... 25:00 2000 1500 1000 500 LOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOL()L()LOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLOLO ...... COLOVet)C\I...--OO')a:> ...... COLOvMC\J...--· ...--C\let)~LOCOf'...CX)O')O...--C\let)vLOCO...... --...--~...--...--...--...--~. I III IIII IIIIIIII Figure. 7.10 a(top)- Histogram'showing number of VSOP observations'.made at relative' wind speeds ranging from 0 to 80 knots in 5 knot intervals. Figure 7.10 b (bottom) - Histogram showing'number ofVSOP observations made.at relative wind directions ranging from -180to 180 0 off bow in 0 10 . intervals l l - 59- l 5000 I. i l 4000 3000 It I. 2000 l 1000 r. f 0 t) I 0 1 2 3 4 5 6 7 8 9 l 0)' 0 -( tt t -.- VSBG8 DIDA • FNCZ. FNEF FNFD • FNGM- i.·.·.·.·.·.·.·.·.·.·.·.·.·...·.·.·.·.·.·.·, r FNGS • ,·.·.·.·.~.·,·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.I , j FNOY • i.·.·f!.~ ••·.·.·.·*·!·.....·~f!;·4·#·.·?· ..·I' •• [ DDLN DDUC • Dm~ - I DHRG • DIMe • DNBR· DNJR • PCEL • 55:: ... . i--i PELT. ~ PELU. MS., .sI, If' PGDG • PGDS. PGDW. PGEG I- I LIMA C6DS : r GBSA • ,·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.1 GBW. GBVW • GJMR .. r11 GJMS • GXES I GZMM • - VRAZ • VSBV3 ~ IBPA .. I Figure 7.lla (top) - Histogram showing number ofVSOP temperature observations rated at exposure 100 indices from 0 (poorly exposed) to 9 (well 80 exposed) 60 40 .~.. : Figure 7.l1b (bottom) - Numbers ofVSOP observations by ship rated at 20_ exposure indices 0 to 9 as in 7.11 a -60 - 4.0 6, (b;d ...... 0 3.5 Medium CD a- Poor ::s 3.0 ...as a- E~posure 4) 2.5 Sensor Q;.e Cl) 2.0 f- a- 1.5 :c 1,.0 Cl> -0" 0 0.5 :s 0.0 a. -0.5 0en > -1.0 0 2 4 6 8 0 1000 4.0 ...-...... ,-,,...-.,~r--'!"'P--r-'l~,...-.,.-.- ...... r-'I~~...... -r-'l~r--""II.. 3.5 3.0 2.5 2.0 1.5 1.0 0.5 0.0 Joo------.=--'If-t---'76Ifo------.(. -0.5 -1.0 ~...-- f,.....~ ----..a---..a- --..a---..__...... o 2 4 6 80 500 1000 4.0 r--...... -...... -,.....--.,,...... -.,...... -.r---...--,.....--.,,.-,r-"""!r---r---r--"I,...... ,,...... ,r--...... -r-'t~~. o ...... 3.5 3.0 2.5 ...... ------.-1····1 2,.0 ... 1.5 < 1.0 0.5 0.0 ....---~..._---__.i~~~....------~ Q.. -0.5 o ~ -1 .0 1.0-"--...... --..--..--'1---'--"-""-1/-&-""-""-""----..-01---'---'-'---'--""----...... 500 1000 Total Cloud Cover (oktas) I Short Wave Radiation (W/m2) Figure7.12a (top) -VSOP measured air temperature differences from model binned on short wave radiation for daytime measurements and total cloud cover for nighttime measurements separately for sensors with good, medium and poor exposures Figure 7.12b (centre) - As 7.12aexcept data divided by measurement teclmique (screen or psychrometer) Figure 7.12c (bottom) - As 1.12a except data divided by relative wind speed at time of measurement l l - 61 - I. ..CD i. :s 3.5 ·····..··0··..··· Poor ...ca I .. _._....._. ~.J1edium t a.CD 3.0 E • Gocxi t CD 2.5 I-... I. .-c 2.0 ...... ········· \ o ··.... I \\ a. 1.5 ... /\f \ l ~ ~-,~ . 1.0 ...... l..... l ...... / \ !. 1 I -f~' · 0.5 0.0 -0.5 Q. o -1.0 1...... ---...-..-...... -..-...... --...... en o 2 4 6 8 0 500 1000 > Total Cloud Cover (oktas) I Short ,Wave Radiation (W/m2) .-.-... 0 ...... 0 1.2 Cl) .. o Psychrometer ..,:s 1.0 ..as .....••..•..•.•.. Screen Cl) 0.8 a. E 0.6 ...... ··1-.. "...-1...... i Cl) .... -.. ..- •... I- 0.4 /J ...... "../ ...... J ...c / z " 0.2 ... 0 ~ , a. I 0.0 ~ -/ ,~ Cl) /'·1....-. C -0.2 1 ---.. CD -0.4 ""C r 0 :s -0.6 -0.8 Q. o en o 2 4 6 8 0 500 1000 I > r Total Cloud Cover (oktas) I Short Wave Radiation (W/m2) r [ Figure 7.13a (top) - VSOP measured dew point temperature differences from model I binned on short wave radiation for daytime measurements and r total cloud cover for nighttime measurements separately for . I sensors with good, medium and poor exposures Figure 7.13b (bottom) - As 7.l3a except data divided by measurement technique (screen orpsychrometer) r ! r 10:(lE}i ·...... -...... ---...... -...... ------.....'i 600a", 50C0:: 4G€Ul 3000 2000 1000 o o LO 0 1.0 0 LO 0 It') C\I C\I C") et) v ~ LO l() o 0' U") 0 Lt') 0 LO 0 to C\I C\I CQ co v -.:;t LO LO ~ N U U ~ ~ ~ ~ o 'r' 9 if .er ~r.. ? tr ..9 'r er ~~g~ _"fi. .. m;·.m·.·.···.m·.·.·.·.~·..... ·.~·.·.·.·.m···.··~·······~········~········rn······rn·· ·m········m········m········rn·····~········~····;···~····....m··..··..~·····m·····..·m·..··..·~·....···%:I···....· PGDW : ,...... :.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:. PGEG ;.;.:.;.;.: ;.;.:.;.:.:.;...;.;.; ;.;.:.;.:.;.:.;.;.:.;.;.;.;.:.;.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.;.:.:.:.:.:.:.;.:.;.:.:.:.:.:.:.:.:.:.:.:.:.;.:.:.:.;.:.:.:.:.;.:.:.:.;.;.:.;:::::::::;:::::::::;:;:::;:::::;::.;.;.;.;.;.:.:...;...;.:.;.;...:.:...;.;.:...:.:.'.;.'.;.:.:.:.:.;.:.' LIMA : :.: ; e60S iiiiiii IIDII&mBDm:;:mE :z;rmm~rnJ5 m~~~~~~~~ ~ ~~~ ·~· ~~:.~~~~~ ····~=·~·.. ~ ····~····· -illi IIILI] ·m·.··.·5.·.··[I··.m··.:! ···5..··..··m..·· :m·..·· rn . - .:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:-:.:.:.:.:.:.:.:.:.;.:.:.:.:.:.:.:.:.:.:.:.:.:-:.:.:.:.:.:.:-:.:.:.:.:.:.:.:.:.:.:.:.:.: GBVW _------u.m::mmmmzm~~.;.;.:.:~.:.;.:.;m.:.:.;.:.:8],J:.:.:.;.;. GJMR -1smllllllBlllllll .mm.mmam:.:.:.;.;m:;.;.;.;.;.:m;.:.:.;.:m.;;.;~.:.;.:.;~.;.:.:.;.~;.;.:.;.;~.:.;.;.:. ~~~::.:.~::.:.~:.~ GJMS~~ ~·.· :;:;mm;:E ·.·~""""~""""~""""~""""'~"''''=''''''''~'''''''''=''''''''~''''[;Z.:•.•:•• ••:..•: •.: .•:.: ::.•.:.:.: ·:::········~· -"1i_mw-iIII------m;...__ ..m.:.·.·.···.m·.·.·.·.m·.·.·.·.m· ...... ·;;z·..··....mJ....··· 1iii~ ~.·~ ~ m.·.·.·.·.·~· ~.·~.··.. mE~~~~~.. VRAZ "" " :: :.; ; :.:.;.: :.;.: ; V!~ <.:.:.".,.:.:.:.:~,.".".: _:::::::: :.:.:"".:.:.:.:.>:.:.:.",.,.:.:.>:.".,.:.:.:.:.»:.".:. :'.:.:.:.:.".:.:.:.:.:.':.:.:. :::::::.:; .:.,.:.:.:.:.:.:.:.:.>:,.:.,.:.:. :.:.:.:.:.,.:.:.:.:.:.,.,.:.,.:.:., .:.:.,.:.".:.:<.,.,.:.:.:.,.:.,. ,.,.:<.:.:.,.,.:.:.:.,.:.,.".:., KR.GJ ;';';':':':":';':';".; : ;.: : ;.:.; :.:.; ;:; :.: . KRJL ;.;.; :;:; ~:~:::::::::::::::::::::::::::::::: ::::::::::::::::::::::::::::::::::: :.;.:.; : :.: :.;.:.:.: .;.: :.: :.:.:.:.: :.:. KRJP . ~B ..:;;; :.:.:.:.:.:.:.:.:-:.:.;.:.;;:.:.:.:.;.:.:.:.;.:.:.:.:.:.:.:.:.:.:.:.:.:. Wt1LG -£i ~:.:.:.m:.:.;.:.:~.:.;.:.;.m;r;.:.;.;.;E.:·:·;·m;r:·:·;·:·;G·:·:·;·:·m:·:·:·:·:fi:·:·:-:.[J;.:.:.:.:3];.:.:.:.:.[J:.:.;.:.:[E.:.:.;·m:·:·:·;·:E;·:·:·:.;·;E:·:·:·:·;Z[:·;·:-;::Z;:·:-:·:·:·§]:·:·:·:·:]];·:·:·:·;·illl:·:·;.:]];·:·;·;.;.§]:-:·:·:·:3];.:.:.:.:.3]:.;.:.:- ·ii·:·:·ii:·:~:·:m·:·:·:·i:·:,:·:i·:·:·:·~:· ~::mmD· ':':Z':':':'E:-:':':~':':':~':':':'~:':- - WPFHZZwp m··m·:m·:·:·:·m:·:·:·;E:·:·:·:m':·:·:·[m:-:·:·:;ID·:·:m·:-:·:·m:·:·:·:E·:·:·:m·:·:·:·[I:·:·:8]':-:':3-:':-:-[8:':':':%':':-:-:8::':-:']]: :%·:·:·:·:8::·:·:·8]:-:·:-:Z·:·:-:-O:·:-:·: Wl:?VF .;.;.;.:.;.:.:.:.;.:.;.: ;.:.:.:. :.:.:.'.:.:.:.:.'.:.:.:.:.:.:.'.:.:.:.:.:.:.:.:.:.:.:.:.:.'.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.'.:.;.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:.:. WPWH :.: :.;.: : : ;.: :::;:;. ~~;:;:;~;:;:;:;:;:;::;:;:;:;:~~~~~.....-r""'r-T.,...,..,...,..,.,..,..,~~r""'rr-rr-r-,.,.,..,..,.,.,..,,.J WSDG ~ ~ ~ ~ ~ ~ ~ ~ 'Po ,..,..,.,..,... l VI m o 0 ~ 0 ~ 0 ~ 0 ~ &0 ~ 0 ModelWind Speed (knots) Figure 8.1 a (top) - Histogram showing number of VSOP observations for model wind speed ranging from 0 to'60 knots 1~~1 Figure 8.1 b (bottom) - Numbers ofVSOP observations by ship for model wind speed ranging from 0 to 60 knots. Shading indicates numbers of observations per 20111 5 knot range. [ ! -63- [ l r t l l 1300 1200 I- -. 1100 I- tn - Cl) 1000 r (,) - - c CD 900 .. I~ .. - ::s 800 (,) .. - I (,) r 0 700 I- - 'I- 600 .. 0 - .. 500 Cl) l .c '"" - E 400 i- - l z::J 300 l 200 100 ~ 0 l I Model Wind Direction (OT) Figure S.lc- Histogram showing number ofVSOP observations made at model wind directions from 0 to 360 °T in 100 intervals. -7(;1 -6®, -50' -40 -3<:1 -20 50 . 40 3.0~~•...••.•.•. ~:g~., 0.0 // 35 -1.0 t I I 30 -70 -60 -50 .-,;.40 -30 -20 -10 o -80 -70 -60 -50 -40 -30 -20 -10 60 55 50 4-5 4-0 18..•..0 •.•..•.•. 16.0~~~' . ~.~0 14.012.0/).••• ..··· •.••····. 10.0 t I I -70 -60 -50 -40 -30 -20 -10 Figure 8.2 (top) - Wind speed difference from model.(knots) averaged in 50 squares over the North. Atlantic. Figure 8.3 (bottom) - Model wind speed (knots) averaged in 50 squares over the North Atlantic. '~'_~---'--"'~---J--~~~~~-~-----\f----··\i---·'j----·~....--~-- ---1llIIo!:~~;"_. ----.-\ VSBG8 i v~~g~... j it· i i r _x "t _ i i t la DIDA K>-t FNCZ FNCZ FNEF IK>-t FNEF FNFD tOt FNFD FNGM I FNGM R\GS ~ R\GS s Ff\OY bW~1 i f f j ) _"V!" "I i ! j I J.. DDLN DDJC DruC l.l-INE I~ CHJE DHOO ... DHFG DIMC DIMC DNBR ~~' DNBR DNJR I DNJR +-- i ~ ~ · i . F' ~ l S i i .. PeEL PeEL PELT PELT PElU PElU FGOO FGOO PGDS PGOS PGDW PGDW R?EG ~. j ; i i. i 1-"::1:: t i i i la llMA C6DS C6DS GBSA GBSA GBW GBW GBWJ GBWJ GJMR GJMR GJMS GJMS GXES GXES m GZMM GZMM VRAZ VRAZ VSBV3 VSBV3 ISPA IBPA I i I I S I 1 I I. !I I I I KRLZ KRLZ KRGJ KRGJ ...... KRJl KRJPKRJL I ~ KRJP KRPB K>-t KRPB WMlG I...-c>--I WMlG WPFl ...... WPFl WPHZ WPHZ WPVF WPVF 'JVP.NH 'JVP.NH \NSfX3 I""U""I""""'I \NSfX3 I" "I." .1". 'i'''' i"" ji". i"-"I .... I .... i. "'i' ... 1.... I 0 1 2 3 4 5 6 7 8 -30 -·25 -20 -15 -10 -5 0 5 10 15 20 25 30 Mean Wind Speed Difference from Model (knots) Mea.n Wind Direction Difference from Model (0) Visual • Figure 8.4a (left)- Mean wind speed difference from model (knots) by 0 Fixed Anemometer method (visually estimated, anemometer, hand held anemometer or unknown) by ship callsign. • Hand Held Anemometer 6 Unknown Figure 8.4b (right) - Mean wind direction difference from model (degrees) by method (visually estimated, anemometer. hand held anemometer or unknown) by ship callsign. 5000 4500 4000 3500 3000 2500 2000 1500 1000 500 o ~Nooom~NOOOID~NON~mooON~WOOON~IDOOON~ NNN-r--r--r--r--r- -r--r-NNNNNMMM IIIII 1 1 f") ~ 1 1 u u (J1 o o (J1 o (J1 o (J1 VSBG8DIDA t===~~:;:~======~~~~~~~~~~+1 FNCZ t·:·:·:·:·:·:+:·:·:·:·:·:·#:·:·:·:·:·:·:I·:·;·:.:.:.:·.:.:.:.:.:.:.'.:.:.:.:.:.:1.:.:.:.:.:.:·,••·.:.:.·.:.:1·:·:·:·:·:·:·.:·:·:·:·:.:·1·:·:·:·:·:·:1·:-:·:·:-:·:·' FNEF 1.·.·.·.·.·.·£·.·······.·.,·.·.·.·.·.·.·,.·.···...... ••..••••••..• .·.·.·.·.·.;.,·.·.·.·.·.·.·1.·.·.·.·.·.;.,.·.·.·.···)·.:.·.·.·.:.:1.···.·.······1:.:.·.·.:.:-3:.: : ).: , FNFD £...... ••.•¥- ,: ,. ············4·············,············:1.·.·.·.";"J FNGM F············j·············I.·.·.·.·.·.·.,·.·.········1············,············,·············1.·.·.·.·.·.·.F············t············j······ ,, FNGS E.·.·.·.·.·.·{.·.....·.·.,·.·.·.·.·.·.·, ,·.·.·.·.·.·.1···········1············,·········.·.·1.·.·····....·1 FNOY ,·:·:·:·:·:·:J·:·:·:·:·:·:·I.·.·.·.·.·.·. ·.·.·,·.·.·.·J:·:·:·:·:·:·:,·:·:·:·:·:·:f:·:·:·:·:·:·j:·:·:·:·:·:·:1 DDLN DDue [·:·:·:·:·:·:1·:·:·:·:·:·:·1:·:·:·:·:·:·:1·:·:.:.;.:.:3.:.:.:.:.:.:.1.:.:-:.:••••,.:.:.:.:.:.:.1:·:·:·:.:·:·:1·:·;.:·:·:.:1:·:·:·:·:·:·1:·:·:·:·:·:·:1·:·:·:·:·:·:·1:·:·:·:·:·:·1-:·:·:·:·:·:1.:.:.:.:.:.:.' £.:.:.:.:.:.:+:.:.:.:.:.:., DHNE £.:.:.:.:.-.;1.:...:.:...:., 1-:-·.:.:.·.:.1:·······:·:1······;·:·:£············i*$·,· ·.·.·.·.·.·.#:.:.·.·.·.·.:I.·.·.:.·.:.·.F:.:.:.·.:.·.I.:.:.:.:.:.:J.:.·,:,:,:,:,' ~ 1:·:·:···:·:·:f.;.:·:·:··.:·1 DHRG ~ F············I············l············,············t· J . ··········.·a··.·.···.···.J·······.···.·j······.·.·.·.I·.·...·.·.·.:I·.·.·.·.·.·.·j.·.·.·.·.·.·., ~ DIMe t···.·.·.·.··j·.·.···.···.·j.·.·.·.·.·.·.,·.·.·.·.·...·E.·.·.·.·.·.· .'.·.·.·.·.·,·.·...·.·.·.1·.·.·.·.·.·.·1.·.·... ·.·.·.,·.·.·.·.·.·.t.·.·.·.·.·.·j.·.·.·.·.·~.J I·:·.·.·.·.·:}·.·.·.·.·.·.I DNBR t·:·:·:·:·:·:)·:·:-:·:·:·:·I:·:·:·:·:·:·:,·:·:·:.:.:.:+:.:.:.:.:.:.*•••••••••••• :·:·:·:·:·:·*·:·:·:·:·:·:1·:·:·:·:·:-:·1:·:·:·:·:·:·:l·;·:·:·:·:·:·'·:·:·:·:-:·:·j:·:·:·:·:·:·:1 DNJR ~ -.!E2iliJ~':':'!:':'::!.'-~I'::':':'i:':':i'I:'i:':':;':':';:I':i':':·i:·:·::..··:····;:·····::::::::::···~··;··~· ••••:i··~···{:·····:*·:i·:·:·;:·:·:;J.:.;:·:·:i·:·:·il:·:~·:·:·;:·:·:i'·:·i:·:·:i·:·:Oj~·:·:i·:·:·;:·:·~j:-:;·:·:·i:·:·:~1 ~r------+ peEL ~ .:.:.:.:.:.;. :.;.:.:.:.:.: .:.:.:.:.:.: .:.;.:.:.:.:..':':':':':': .:.'.:.:.:.:. :.:.:.:.:.'.:. .:.'.:.'.:.:. :.:.:...:.'.: .:.:.:.:.:.;. :.:.:.:.:.:.' .:.:.:.:.:.: .:.:.:.:.:.;. :.:.;.:.:.:.: ';':'.':':':":':':':':':' :.....:.:...:...:...:..... PELT t.·.·.·.:.:.:.F".·.:.·.·.·.J:.·.·.·.·.·.:I.:.·.·.·.·.:.f·.·.·.·.·.·.:g.·.·.·.·.·.·j.·.·.·.·.·.·.,·.·.·.~.·.·.·I.·.·.:.:.·.:.F: ...·.·.·.:.:t.:.·.·.·.·.·f 1:.:.·.:.:.:.:t···:·:·:·:·:l:·:·····:·:·J PELU ~ F:.;.···.:.·.I:·:·····:···:J.··:·········F:·:·········:J:,···········4·············,·············,.·.·.·.·.·.·.I:.·.:.:.:.·.I.:.:.:.:.:.:J.:.:.·.:,'·:·':"':':':':':£':':':"·:·:1 ~ PGDG ,.·.·.·:·:·:-:1·:·.·.· ·:·£:·:·.·:·.·.·*·.·.·.·.·.·. • •• .·$1$·.,·.·.·.·.·.·.·1.·.·.·.·.·.·.,·:·:·:·:·.·:£.·:·..: j: ,.: -:.;.:., ~ PGOS ~ ....·.·:·.·.·;I·.·.·.·:·.·.·f.·.·.·.·.·.·1'.·.·.·.·.·.£.·...... ··w· ·.·.·.·.....·I.-.·.·.....·.(·.·;·.·.·.·.·j·.·.·.·.·.·.·I.·.·.·.·.·.·.,·.·.·.·.·.·.·f.·.·:·.·:·.·.I.·:·:·.·.·.1 ~ PGOW ~ £:·:·:·:·:·:-1·:·:·;.:·:·:1·:·:·:·:·:·:·,.·.·.·.·.·.·.,·.·.·.·.·.·.$.·.·.·.·....*·.·.·.·.·.·.,·:.;·:·:·:·:·1:·:·:·:·:·:·t:·:·:·:,;·l:·;.:·:·:·:·j:·:·:·:·:·:·:'·;.:·;·:.;·:·j CZJ ~ PGEG F:·:·:·:·:·:·I:·:·:·:·:·:·:I·:-:·:·:·:·:·':·:·:·:.:.:.1':':':':':':3':':':':':':' • •••&k§:.:.:.:.:.:.:l.:.:.:.:o:.:f:·:·:·:.:.:·j:.:.:.:.:.:.:1·:·:.:·:.:·:·1:·:·:·:·:·:·1·:·:·:·:·:·:J·:.:.:·:.:·:·I:-:·:·:·:·:·:t·:·:·:·:·:·:J LIMA e60S GBSA ,·.·.·.·.·.·.·I.·.· ·.·.·.F· ·······I········ j , ,..; . ··SS .·.·.·.·.·.·...·.·.·.·.·.·.j·.·...·.·...·I.·.·.·.·.····,···········t··....······j··· ·······,·············£······..····1 GBW f...·...·.·..,·...... ·.·.j 1.....·.·.·...,·...·.....·.:£.·.·.·.·.·.·+.$ •. .·.·.·.·.·.·;·.·.·.·.·.·.,·.·.·.·.·.·.·1.·.·.·.·.·.·.,..·.·.·.····1;..·······.·1.·.·:·.·.···:1 mm GBVW E2iliJ 1·:·:·:·:·;·;.I:·;·:·:·:·:·:I·;·:·:·:·:·:·E··:········· ·:·:·:·:·:·;·f:·:·:·:·:·:·*·:·:·:·:·:·;I·:-:·:·:.:.;};.:.:.:.:{;.:.:.:.:.:) GJMR t.:.:.:.:.:.:I.:.:.:.:.:.:.' ~E.:.~:.:.:];.:.;.~F•••~.:•••;;;,:;:••• IIllllIlI.__••••_IlllIlIE:...~:.."~'4~....~.....~;..iI].:.:.~:.;.:~.:.IC:':"~.. :'::8l.:.;I[;':':];':':'B3:.:.j];.;.:.~:.:.::;].:.j~:-:':];';';':3:·:1 f8]]J GJMS F:·:·:·:·:·:·j:·:·;·:·:·:·:I·:·:·:·:·:·:·E:·;·:·:-:·:·*·:·:·:·:·:·:4·:·:·:·:·:·;·':·:·:·:·:·:·:'.:.:.:.:.:.:-F:.:.:.:.:.:. } g•• :.:.:.:•••:.:1:.: 0 : 0 :.:.:.:1.:.:.:.:.:.:.1:.:.:.:.:·:·:1·:·:·:·:·:-:+:·:·:·:·:·:·1:·:·:·:·:·:·:1·:-:':':':':'F:':':':':':'~ GXES t.:.:.·.·····j····.·.···.;.I:.;.:.:.:.:.·£.···.·.·...:1'.•...... {{4············4·············,:·······:·:.;(·:·:·····:4:·:·······:·);···:·:·:·;.:1 ~ GZMM ,·············£······..····,············1·······•.•.. ············1············4·············1·.·...·.·...·E.·.·...·.·.·.,.·.·.·.·.·.·.J·.·.·..···1 VRAZ 1·············,············1···········;.··:····.:..£ ...... •....., $ ..: j: :.•.: ,···:···:····t·····:·:····I:·.·.·.·..;·:1 ~ VSBV3 ··········4·:·:·:·:·:·:1-:-:·:···:·,.£:·:·:···:···*·:.:.:.:.:.:J ~ ~ IBPA :.:.:.:.:.:.:.:.:.:.:.:.:.:.:-:.:.:-:.:.:.:.: : : : : . KRLZ •••• ••:{{{.:J{-:+¥$.i:.:.:.:.:.:.:1·:·:·:·:·:·:3·:·:·:·:·:·:·1:·:·:·:·:-:·:1-:·:.:.:.:.:·\;.:·:·:·:·:·:1:-:·:·:·:·:·:I·:·:·:·:·:·:·j KRGJ KRJL F:.:·:·······f:.:.:.·.:.:.:I.:.:.:.:.·.·.F:.·.······:·1·····:·····:£············,············: ? •••••.•••.•••.•.•~ .....:.....:.J:.....:...:-:l...:.:.;.:...F:-··;·:.;·:·:I:-:·:·;·:·:4:-:.;·:··,:,) Ern ~ KRJP 1.·.·.·.·.· ,·.· · 1.·.·.·.·.·.·1.·.·.·.·.·.·.,••·.·.·.·.·.·i.·.·.·.·.·.·¥· ·.·.·.·.j·.·.·.·.·.·.·,#·%:...;;:1·.·.·.·.·.·.·1:··· ·····1·····.·.·~·.i·.·.· ·I.· · ·.·.l·.·.·.·.·.·.1·: j ,·.·.·:·.·:·:·1.·.·.·.·.·.·:,·.·.·.·.·.·.3·:·.·.·.·.·.·,·.·.·.·.·.·.61\,$ ...• ••• ~·.·.·.·.·.·.·.·.·I.·,·.·.·.·.·:t.·.·.· ..·.·.-:t·.·:·.·.·:·.·1:··.·.·.·..<·:1·.·.·.·.·.·:.,·.·.·.·.·4·:·j:·:·.·:·.·.·.1 KRPB E:·~·:·:·:·:·:t·:·:·:·:·:·:t·:·:·:·:·:·:·I:·:.:.:·:·:·:,·:·:·:·:·:·:·£:·:·:·:·:-:·j:·:·:·:·:·:·;I·;·:·:'·:·:·:·1:·:·.·.·.·.·*·.·.· 1:.:.:.:.:.:.1.:.: 4 :.:.:.:1.:.:.:.:.:.:.1:·:·:·:·:·:·:t·:·:·:·;·:·:·I·:·:·:·:·:·:·j:·:~:·:·:·:·:1 WMLG ...·.·:3·.·.·:·.·:·:·1.·:·.·.·.·.·:,.:••.•••••..• WPFZ t-:.:.:.:.:.:J.:.:.:.:·:·:-J:.:.:.:.:.:.:t.:.:.:·:·:·:+:·:·:·:·;·:·1;·:·:·:·:·:·:1.:.:.:.:.:.:•••• ••••••••••.•••:.:.:.:.:.:.1:·:·:·:·:.:·1·:·:·:·:·;·:1·:-:·:·:.:.:-1:·:.:-:.:-:.:£.:.:.:-:.;-:+:.:-:.:.:.:., WPHZ '·:·:·:·:·:·:·1:·:·:·:·:·:·:1·:·:·:·:·:·:·'·:·:·:·:·:·:·t·:·:·:·:·:·:I.:.:.:.:.:.:.f ••••• :••• g...... •••• $,:.:.:.:.:.:.1:.:.:.:.:.:.:(.:.:-:.:.:.;+:-:·:·:·:-:·I:·:·:·:·:·:·:I·:·:·:·:·:·:·F:·:·:·:·:·:·:1:·:·:·:·:·:·:1·:·:-:·:·:·:·' WPVF 1.:.:.:.:.:.:..11·:·:···:·:·:&·:·:·:·:···:1:·····:.:.:.1:.:.:.:.:.:·;1.·.·.:.:.·.·.1:.:.:.·.:.:.: WPWH WSDG 1 1 1 o (]I u l\.) ~ .1- O'i o (J1 o o VSOP - Model Wind Speed (knots) Figure a.5a (top) - Histogram showing number of VSOP observations differing from model wind speeds ranging from -25 to 35 knots Figure 8.5b (bottom) - Numbers ofVSOP observations differing from model wind speeds by ship for wind speed differences ranging from -25 to 35 QC. Shading indicates numbers of observations per 2 knot range. t l - 61- 1.1 I 1.0 0.9 o Visual l 0.8 I 0.7 Measured t 0.6 • I 0.5 0.4 r 0.3 0.2 l 0.1 t 1.0 -90 -70 -30-50 -1 0 10 30 50 70 90 I I Oi o 8 VSBG81=:!!~::;~~~======~~~~~~:;~~tDIDA ~ FNCZ FNEF FNFD FNGM FNGS f FNOY DOLN DDUC r DHNE I" DHRG DIMe DNBR rmm DpNCEJR 1-~~;r~~~m:m~:.:.:~. -:~=~=====:":"§:':"'=:':'¥'::':':'::f:':'::':'¥'::':':'::'f:':'::':':"~:':':':'~f:.:.:~.:.:,.~:.:.:.:.~, -:~~-~mr-r- L , .. .:.:':.:.:.:.:.:.:.:.:.:.:...... :..••..• :.:.:.:.:.:.:.:....:.:.:.::...:.:..:.:.:.:.:.:.:.:.: l PELT PELU PGDG PGDS PGDW PGEG LIMA C60S r GBSA GBvVGBVW GJMR GJMS GXES GZMM VRAZ VSBV3IBPA I VSOP - Model Wind Direction ( 0 ) Figure a.Bc (top) - Histogram showing number of VSOP observations differing from model wind direction ranging from -90° to 90° for visually 100 estimated and anemometer wind directions 80 60 Figure a.5d (bottom) - Numbers ofVSOP observations (all methods) differing from model wind direction by ship ~g:.t for wind direction differences ranging from 200° to 200°. Shading indicates numbers of observations per 10° range. 60 ...... 'I"'P"I"'I.....I"'I""'I"'IP"I""'I"'t~...... r""'r"I"""I...... r"'f""I"'I...... ,...... ,....,...... ,."'I""I"""I...... '"I""I""I''''I'"WI 55; 50, 45 40, 35 3-0 25 20 15 to 5" o "-I.oI.o ~ ~ ~ " o 5 t,·O, 1",·5· 20i 2,,5 30 3.5 4·0. 45. 5,0, 55 6'{); 10 ...... 'I'"'I"""I"...... 'I'"'I"""I"r"r"'I""I~I""I"""P'I~...... ""I'""I"'9...... ~...... "'PP"I"...... "I""'I"I 9 8, {' 6 5 4 3.:· 2 r O: ~ ~"'· (1 5 10:1520 2'5< 3,:0: 3':5:' 4·0 4/5 5:0'5:5 6.0 VSOp'measuredwindspeed,"(knots)binned_on modelwind speed,(knots) Fimrre.8.6b. (bottom) Differences in VSOP measured wind, speed.from, model'values (knOlS} binned on modelwind speed (knots) - 69- o • VSOP 50 100 150 200 250 300 350 Model Wind Direction (OT) 5.0 ----. ..In 0 c 4.5 ~ ...... " -c 4.0 Cl) Cl) Q. en 3.5 -c c i 3.0 Q) 2.5 -0 0 :E 2.0 Q. 0 1.5 en > 1.0 r 0 50 100 150 200 250 300 350 Model Wind Direction (OT) t Figure 8.6 c (top) - VSOP and model wind.speeds (knots) birmed on model wind r direction (OT) f Figure 8.6 d (bottom) - Differences in VSOP measured wind speeds from model values (knots) birmed on model wind direction (0T) r r I r I r ( I I - Iel:. . ' t 0 1"T"T.,..,...,l"TTrI'""'I.,....,..I""T"'T'.,.....,I"'T"'I""TT"'T-ror-I"'I""T"'r'r'1""'r"""1""...-T-Y""I""'I"""II"""r"""I'~I"""or-r'...... "" 9 8 ,~.'. . \ 7 r ....------"""II ;, . "'\ ,. . \ 6 ;; \ . 5, l ,., 4 ,. 3~ 2' O...... ,..t....A."""-I.o..&..A-oIo...... -...... -'-i.o...r...a-L.;,I....Io,...... ~""""""""...... ·•··· o 5 10 15 20 25 30 35 40 45 50 Model Wind Speed (knots.) Figure 8.Ta.(top) - VSOPmeasuredwind speeds (knots) binned on model wind speed (knots) separately for fixed anemometers, hand held anemometers' and visually estimated winds. Figure 8.7b (centre) - VSOP measured wind speeddifferences from model (knots) binned on model wind speed (knots) for sensors as in 8.7a Figure 8.7e (bottom) - As 8.7b except datadivided by visually estimated and measured wind speeds and also bystable (air temperature greater than sea temperature) and unstable conditions (air temperature less.than sea temperature) Figure 8.8 a (top) VSOP measured.wind speed (knots) binned on analysis model wind speed (knots) by country for anemometer measured data. Ocean Weather Ship Cumulus wind speed data are also plotted against forecast model wind speeds. Figure 8.8 b (bottom) Differences in VSOP and Ocean Weather Ship Cumulus measured wind speed (knots) from analysis model values binned on model wind speed (knots). Differences in Ocean Weather Ship Cumulus wind speeds from forecast model data are also plotted. - 72 - ~ ..en 0 c ~ ...... 10 G) "C 9 0 :E 8 E 0 I- 7 "l- i Q) 6 (,) t: i Cl.) ~ l- 5 ....Q) ...... 4 i C t "'C 3 Cl) Cl) c. 2 il ! (/J "'C 1 •••••~.~ ••~••••~•••M...... •••••••••••••.. C ~ 0 c 0 5 1 0 1 5 20 25 30 35 40 45 50 ca (1) :E Height of Anemometer asl (m) Figure 8.9- Graph showing VSOP wind speed difference from model (knots) by ship (scatter points). Line indicates· correction from anemometer height to 10 m above sea level. l - 73- 5000 l 4000 l 3000 2000 1000 o o 1 2 3 4 5 6 7 8 9 o 7 ~ ~ t ~ ~ ~ ~ ~ .... o VSBG8DIDA --r~~~~~~==~"~""""~""""~""""~""""~""""~~!!::::~~::::~~~~==~:;:~, . FNCZ I: FNEF FNFD· i~f.!··f!.···:e!~*·····:······~·····:·5 FNGM • FNGS • , , FNOY • E························:·················, DDLN ..------~----~------=====:L DDUC ~ DHNE DHRG: DIMe. DNBR. DNJRPCEL .,.r------I- ~~. ~ ~w. ~ J?GDG • ~ ffi PGDS. z PGOW • ~ J?GEG .. ~ LIMA -- ~ CGOS : '"" GBSA. GBW. GBVW • GJMR • GJMS GXES • GZMM • VRAZ • VSBV3IBPA·•...------.:..:.:.:..:::::::::::==.------======1. KRLZ KRGJ • KRJL- KRJP- KRPB WMLG WPFZ • WPHZ • Wl?VF WPWH : WSDG ~r-r-rr.,..,.-r-,r-rTT-r-I...,.-r-r-r-T-r-1r-r-rT.,...... ,..,..,...,...... ,~ .....!!!!p!!!!IlFlI!I!I!pBIip!lllllll...... IIIIIIpIIl,..;;;;;;;p;;;;;;&.. o ~ ~ ~ l & ~ ~ ~ ~ Figure 8.1 Oa (top) - Histogram showing number of VSOP wind observations rated at exposure indices from 0 (poorly exposed) to 9 (well exposed) Figure 8.l0b (bottom) - Numbers ofVSOP observations by ship rated at exposure indices 0 to 9 as in 8.1 Oa 60 ~ "I""I""'tP""P'l"'."...... ~~r-'I""""f'I' 55, 50 45 40 35 30 25 20 D. o 15 en > 10 5 O ...... ~~...... ~Ii..IioooI ...... A..Ii.ioIII.Ii-Io...... ~...... ~...... o 5 10 15, 2025 30 3540 45 50 Model' Wind Speed' (knots) "10 ...... !""I""'P"...... r"'P'P...... ,...... ~...... I"""'I""'I'~r"'P"t!""l""'P"...... r-I""!I"~r-t""I, 9 Good Medrum 8 5 , '.. 4 ~ 3 ...... 1...... 1./ • 2 a., 0' 1 UJ > ') ~ : o 5 1 0 15 20 25 30 3,5 40 45 50 Figure 8.11 a (top) VSOP measured wind. speed (knots) binned on model wind speed (knots) by anemometer exposure. Good represents anemometer exposure indices 6 to'9 andmedlum represents anemometer exposures 3 to 5. Figure ·8. 11 b (bottom) Differences in VSOP measured wind speeds from model values {knots) binned on model wind speeds {knots) by anemometer exposure as in 8.11 a. - 75- r 100 I 90 80 ~ 70 0 G) 60 > ;: .!! 50 :a E 40 I :a 0 30" 20 10 O ~ ,...",...... 05101520253035404550556065707580 I Reported - Calculated I Relative Wind Speed (knots) I f 95 90 85 o~ G) 80 .~ ..ca 75 ::s E 70 ::s o 65 60 55 50 ~...... --....----a._~..a.-..a..- ...... --,---...--...... ",,",-.a.-..a--...... -...... o 20 40 60 80 100 120 140 160 180 Reported - Calculated I Relative Wind Direction (0) t Figure 8. 12a (top) - Cumulative percentage plot of difference in relative wind speed (knots) reported by VSOP and that calculated from reported true wind speed and direction and ships speed and heading. Figure 8.12b (bottom) - Cumulative percentage plot of difference in relative wind direction (0) reported by VSOP and that calculated from reported true wind speed and direction and ships speed and heading, -76 ... 60.....-. I"'¥""'I"' ,...... ,..,...... "'I""'I"""I""I"""II' I"""'I""'I"P""r"I ~...... 55 G'ermany ,50 Netherlands UK 4'5 USA 40, 85 ,30 25 20, 15 :10 5, O...... ,..a....r...~I.o&.ooI. ~~ ~ .&.&...I .... o 50 Model Wind 'Speed .(kn,ots) to .P""I"""'I...... ,...... -rI~...... ~...... ,~...... I"'II"-P'.... 9 ··..·····,;0·..····· Germany < ...... Netherlands 8 ...._.....- UK ...... -0.... USA ..... 7 6 5 4 3 .2' ",1 O...... Iio;"I"",I,;...... I.o.I..I.~...... ~...&.oI...... I.ioIiioIoi.~...... ~.... o ,5 '10 152025 30 3'540 Model Wind '.Spe:ed.(knots) Figure8.13a{tQP)- VSOP measuredwind$peed(knots)'binned on model wind $peed (knots)bycoUlltryfor visUally estimated data. Figure.8.13b (bottom)- Differences in VSOPmeasured wind speed (knots) from model values'binned.on model wind speed .• ·(knots). - 77 - I: 10 Pi'="'i~;porjiiiij"""""""'P-W-P...... -tP"P9""'''''''''''''''''''''''''I'''""P'P~''''''''''''''''''''''''''''' 9 Anemometer Day I • Anemometer t-~ight 8 _._.-0-._. Visual Day -c _._ ....._. Visual Night CD 7 c.CD en 6 -c c i 5 4 3 2 oQ. en 1 > O ~ ..I.o.Ioo,j~ .....",.. ~~ ..... o 5 10 15 20 25 30 35 40 45 50 [ Model Wind Speed (knots) l 10 ~ ~ I""'9""P~ ,...... 9 Visual Day l • Visual Night 8 _._.-0-._. Vis (Fix) Day '"0 _._ ....._. Vis (Fix) Night CD CD 7 C. UJ 6 '"0 C i 5 4 CD '"0 o 3 :E 2 Q. o 1 (/J ~ > O ~ ...... o 5 10 15 20 25 30 35 40 45 50 t Model Wind Speed (knots) I Figure 8.l4a (top) - VSOP reported wind speed (knots) binned on model wind speed (knots) separately for anemometer and visual winds and for day and night observations. Figure 8.l4b (bottom) - VSOP measured wind speed (knots) binned on model wind speed (knots) separately for visual winds· reported on ships with and without fixed anemometers and for day and night observations. o VSBG8,; ...... -1.-0 DIDA .. FNCZ...... FNEF' .. 'J FNFl?_ FNGM. _ FNGS .... ", FNOY \" :.: .'.: .'. '.:. '.' '..'. '.' :.' .' '.' ' :.' ' '.' :.: .' :.: '.' .:.: .: '.' ~.. C t·; ·l;·:I.·.:t·;l···'. ',:,{·.1:,·,J:··1 .' ~ ...... •.•.•...... •••...... •••.....•...... ~~.E:J.~ ...... ·lllIf···'····,···;···J···±··".··1 m ~ rn m .- ~ :.:.IQOC•••• ':'-":':+;'•.:..:. :.:010:.: .:••':' :.:. :·;8[2L[;1] ~ rn I]] [J. [:] ~ .:..:.:.:.:.;.:.; ... ;.:.;.:.:.:.:...;.:.::.:.;...:.;.:.;.: ~ .. ~ p.. ~·m.· :5 :.: ;.: .:..':' :.:. .:..':' :.:. :.: .:.: :.:. :.; .:.: .:, .':' :.: :.; £8EJ rn Cl Q ... .'. :.' : : :···········:···:···:·:·······:···:···.rn.~ ~ ...... ""' k Cl m. m f.·.1:···':4:·-J.:·r as(;;D< 'C' I' .. II I 0 J. ~~" u 1. J ~ CO"', 0 0 0 0 0:. O' o VSOP MeasuredWind··.Spee,d,(krrots) Figure~8.15a(top)... - Histogram.showmg,numberofestirnated VSOP .wind speed observations for VSOP wind 100 speed ranging from 080 knots. Numbers 80 indicate Beaufort Scale· Intervals. 60 Figure 8.15b (bottom) - NurnbersofestimatedVSOP wind speed ~g. observations byshipforVSOP windspeed ranging from 0.to 80 knots .Shading indicates numbers of observations per '.'1 knot"range; - 79- 10 ...... ,...... ,,...... ,..-.- ..... ~ * Code 1100 .!!o 9 c ····C···· Kaufeld ~ 8 ········6······· CMM i 7 --0-- Fixed anemometers Q) c. 0, en 6 " " -0 6. '0·····0···.. c 5 ...... 4 2 a. o en > 0 1.oAo.oI~ .... o 5 10 15 20 25 30 35 40 Model wind .speed (knots) ~ (I) 0··' 6 ...-....-...... -...-...- ...-.....- ...-....- -.. c ~ ...... "'C Q) Q) 4 c. en "'C c 2 " 3: ," •••••..[1••• ",0------..------...--~------O...... 0' ,...... ,,'" .a . . .. c' -. ...Q) -2 Q) ....a"'" " E a--...... Kaufeld o E Q) -4 r..- .a.-....-~~~~..a._._..._...... -.-...... -.....--.....-...... c < o 10 20 30 40 Model Wind Speed (knots) Figure 8.16a (top) - Differences in visually estimated VSOP and model wind speeds with different Beaufort Scale conversions ( Code 1100, Kaufeld and CNlM) andVSOP anemometer winds against model wind speed (knots). Figure 8.16b (bottom) - Differences in VSOP anemometer and visually estimated winds against model wind speed (knots) for Beaufort Scale conversions as 8.16a. -00 - 10000 9000 8000 7000" 6.000 50'00 4Q(J(J 30flO 2()()Q 10€lO Q '~ '~.' ~~, ·.·.~I '.'.it,••••'0'..'0'0 It••••••••••tototoj '0·.·.·.·0 •.t.to to.!to.....• t. ;,,,',tit ,'.·..to •• ·Eoto't'.',;',t, to't't ,...:...."....•.:...:..•:.:.:.... ~:.:.:. :.:.:.:.: :.:.:.: ' : t.·I·.·I·.·.·.·.·I·.·~·I·.·.·t·'·I·.·.· r.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.,..·.·t.·.·.·.·.·.·········.·.·····.···.·.···,··.·.:.·.i.···.~.· ..••.•§•••;.••.•... _~ •••.••••••••••••••.••••• ~•••••••••..•J r.·.·.·.·.·.·.·.·.·.·.·.·.~.·.·..·.·.·.·t.·.!'.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·,·1.·.:.·.·.·.·.·.·.·.·.·' ~.•..~¥ ? ~S S$ $ ., 9.·. .t. *-s' E.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·f.·.·.·.·.· ~ '..,•...... •;..•••..•....•..•...•.••l¥..• :•.•...... •.•..... ~.•...•...•..•.... :...... •,..: :...•.:. :.:..•....•.:..•...•: '..•.: :.:. .,.; : :: :.~-:.;....•.•...:. ·8& F·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.·.,·.·.'.·.·.·.•....•.•...... •...•.•..." S $·$*W· '-~·~.f······~····;~·~·~······~·······i ~ - ,...... ••.....•.•••. ;...•...... _•••••••••••••1••••••••• ••••••••'!'•••-J - ~ ~t ~., ,;§.~ §.~ '* ,· tli••••••t ••• •• ••, ••••••••••••••• ••• f ?" }} ·.ill ... I.·.·.~~·.·.·.·.··.·.·.'.·.·.·.·.·.·.·.·*·.·.·.·.· ~ .•..•.* I ••••, ••••••• ·.·.·.·.·...... ·.·.·.·.·.·.·.·.·.·.1 .. ~ f.·~·.·.·!~.·.·~·.~.·.·.·.·.·,;·.·.·I·t',.·.·.·,;·.·... ·,.I·.·II.·.·.·.·.·.·.·.~I~··.·.·.·.·.·.·.·.·.·.·.·.·.·.-·.·.· ...·IW·; ',. & ...... t •••••••••j . t:·••...'.': :•.•:.: :.: :.:.f:.:.:'.':'.':.: :~., :"':" .:.:.t·$••••••••••••~•••••, ••~$ ~ ~••••••• ,:..•:•••:.:.:.;•••:•..:•.•:•.•••:•••;1 f:.:.:.:.:.:.·.:.:.·.:.:.~.:.:.:.:.:.:·J:·:·:.: •.·:,:.:.:.:.:.:.:.•. :.:.:.:.:~ .. ModelPressure (mb) Figure 9.la (top) - Histogram.showing number ofVSOP observations farmadel pressures ranging from 940 to l050 fib Figure 9.1b (bottom) - Numbers ofVSOP observations by ship for model pressures ranging from 940 to 1050 rob. Shading indicates numbers of observations per 10 rob range. - 81 - -80 -70 -60 -50 -40 -30 -20 -10 o 60 ~~-'--I~r...... L.-a.--.-"""""'''''''''''''"",~~~~~~~~rnr60 55 55 50 45 45 4-0 40 o.5~0.0 ~ -O.5~ -1.0 / 35 35 -1.5, I I 30~~~~~~~~....-4-~~"""'~~~~""""""~iIl-WI -80 -70 -60 -50 -40 -30 -20 -10 Figure 9.2 - Pressure difference from model (rob) averaged in 5° squares over the North Altantic. 0.10 .c."...... 0.05 E """'--" 0.00 ...(I) ::J -0.05 0 0 ...CD -0.10 Q. -0.15 CD "D 0 -0.20 :is -0.25 Q. -0.30 0 (fJ > -0.35 -0.40 - 5 0 5 1 0 1 5 20 25 30 VSOP Air Temperature (OC) Figure 9.3 - Differences in VSOP measured pressures from model values (rob) binned on VSOP measured air temperature (0G) - 82- USBG8 ...... ------+---+------+--...-...... DIOR FNCZ ~j::: FNEFFNFD • ~:::~ FNGSFNGM .• t~ FNOY tOt ' ~ DDLN ..------t~-~...... _-~--~----too DDUCD~E H:>-I. ~:::::. ~ DHRG· ~:::'" K.>t:..::::.- . DIMCDNBR t-6-4 DNJR ... +-__~-__I--~~-_._+~-.__..... peEL P'ELT PELU PGOG PGDS PGDW PGEG, L.IMA -1------+------.------+ C6'D-S HIt ~ g~ij3 ~.~::::: OBVW -. !MII·• ...., .•..... GJMR GJMS GXES ~ ~ GZMM ~ URAZ j: MIM VSBU3 e tBPA ..---...... --~-----+----+----+-----+- KRLZ KAGJ ~ KRJL ~ KRJP . t-O-!4 KRPB' jt- • PAS o A-Illllogue Mean Pressu.re Difference .from- Mod.el (m'b) Figure 9.4- Mean pressure difference from model. (rob). by ship callsigp.. - 83- 8000 7000 6000 5000 4000 3000 2000 . 1000 o Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt) Lt). Lt) Lt) Lt) Lt) Lt) Lt) Lt) 10 Lt) Lt) Lt) Lt) Lt) ~MN~cim~~~~~~N~cici~N~~~w~oomci~N~~ ~~~~~I IIIIIIIII ~~~~~ IIIII I N o t::::=:::::;;:~~::!~====~t:~!t~=::::=~~~=t VSBG8DIDA :1 FNCZ FNEF m * ,, * ,%... .·.·.·.·1.·.·.·,·, m r FNFD m t.·.·.·.J·.·.·.·... .,, ~ FNGM ml m la E:l:m ,, •• 'E'" ., , I FNGS B ,········,··s···I·······j········j F FNOY l-----...-,m-mmm.••:••:-.i.:•••••.: __:.••:••:.I:·:···:·, m=-m--~------+ DOLN m DDUC _ .·.·.·.:I···-····,·····:··,·······:I·&:··W···>:·····:·, I m:J ,·.·.·.·.F·.·.·.·,·.·.·.·,·.·.·.·j.·.·.·.·,....•.•.,.g*,. ··2···'····g·,·······,·······, 1.·.·.·.·I.·.·.·.·j DHNE DHRG tm , 1•••••••4 EF·····;m··!··m1·····'m··n·m·p·m·..__.·aME·····m;···'··m..···j DIMe mI , ·., I·······,·······¥·,..t... .::;.%:,..•.•••.,••••.•••,••.••••.,.•••••.! ,·.·.·.·!.·.·.·.·I DNBR la Em , •••••••1:•••••••1:•••••••1:·····:·1·8··· . ·~i·:·:···:,·······*···:·:·j DNCEJRL f------..!mm!.....:~--~£.:...... :g:,$====';':':":I'!:"':':~I.:.:~...:t'~;'M~c·~:·:·:·~4:·:·;~.;.I:~·:·:·:·~';·;·~:·:·E;~.;.;.;~.E;.;.~;.;f~:.:.:.:~J ~~ I .:.:.:.:.:.: :.:.:.: . P , PELT ,••••••••,...... •••••••• j.••••••., •••.••••, m::2 PELU 1i.·.·.·.,·.·.·.·.,·.·.·.·.,·.·&.·••·.·.;1-'i.·.·.·.·j.·.·.·.·I.·.·.·.·, PGDG E·······t·······I..·...·.I·.....·.,·.·.·... ••••%••j , PGDS la ml······Sfl··'···m....·__m····;m··j···m·····' ~ PGDW e F:..·:·;I·.·:.:·;I.:...:.;I.: ;.:I.;.:.... lm3 i PGEG 1------,····.---;mfl..•:.·.:.;;m:.J.:.:.9:.4.:;g··:::::.::.·.a::';':'·"r;mfim;:;r·':' -1- LIMA CGDS rm F·.·.·.·,·.·.·.·,.·.·.·.·j.·.·.·.·,.·.·.·.· .••..•..,,..-.-.-! GBSA ml······m··'···m!·s··,!m······3!l··__~·····m···I··mJ······, ~ Em GBW ,..••.•••,.•....., •••..••, •••••.•1..•.•·.·,.·.·.·.·w·.· ·.·.·.·.1·.·.·.·.1·.••·.·;, ~ GBVW m mm Bc····Bl···,;·m....:· ~···:··m·:'···m....·1 GJMR 6-'.:.:«.·.:.;.;1.·.·.:., mc·;··9·:·j :·m:·····m,.···IB···I;m;.:.··:·..__m·!·:·.*Wm:J·.·:,.j "·:·;·;·1:·;·',:,' Em;] GJMS m mc····~···J:·m··:·:·GJ··:·;B···i...... _lIIIm···:.m;:.:I·mJ··;·;·:1 GXES . liD em mE§··DH-·m······J!N·g···g·:mt···..__m·····m;···'··m······, rim GZMM m ~E····D···'··m······'m·····D···'Ett. m·······m;·'····;m····'m·······:;nj·f····§···tm...... •J VRAZ Em BE····D···'··m;······'BMD§·__m·······m·I.·.·~.•.·, im m m VSBV3 L --2cm:J~········!.__~m~.·.~I::i:..:==:.~s~i~..I ~f~•••••••f~...•.~.•f...~J ~~~•••• _~ 1BpA 1 :.: :.:.:.:.: ::. KRLZ KRGJ Em m'·····~···i··~······mR·:·:D:·· _.··m····;·,G·····~···I··m······m'·····a;··.F··m·····! E.·.·.·.:t.·.·.·.j ,········,·······4 E·······,········,········,········ ,, EiJ , j .:I t:m KRJL E.·.·.·.J·.·.·.·.,·.·.·.·.,·.·.·.·~ ~ KRJP m'······m··j···m·····1m....··mJ··E.·.m!·.·.·__e···.·D.·.'·.m······, I·.·.·.·.,·.·.·.·., mJ m'·····m··j.··D·····j~.....·m··I···m·····__~··m·.·.··JIm·....·m··J·.·m....:t E.·.·.·.·E.·.·.·.i KRPB m mE 1:""':'4:':';';'4:·:';';'£:':';';' .:.:.••:f.:.;.;.:'.;.:.:.:I.:.:.:.¥-:.:.:.J. m WMLG ;~.:.: :.~;.;.:. _m·@i!@··m.·:.:·:m'·:·:·m;:·:'·m··:.:.:m'·:.·§.:.*.~:.: WPFZ -+:O:O:.:.;.;:ozo:o••;oo:":"J:;." •••T.':"::.:.:i':'r.'.;. F·.·.·.·,.:.:.·.:I.:.:.·.·,.:.·.:.:I.:.:.·., (i§··:a·:·J:·;m:·····mi:.·.·g.:... ••.j 1:.:.:·:·1:.:.:.:.•:.:.:.:.' Em . WPHZ m' ~f":':~':+:'~:':':+~:':':"""'J':'''''''''':''' Bm Ii!t·:·:B·:·j:m·:····· ·;E?:.:mC'·:·:·m;:·:I.~:·:·:·:m'···:·n1:·i ·m:·:·:·' 1:,:,:,:,".:·:·:·1:.:.:.:·1:.:.":'( ~ m WPVF ImJ Em F·······,········, ,.·.·.·.·1·······%·······,·······,·······1·...... ·.s·ij.·.·.·.·I.·.·.·.·,.·.·.·.·t.·.·.·.j I:.·.·.·.,·.·.·.·., WPWH a EE····a;···'··;m······mt.·.·;;;J.·.tm..os··__.··m······EJ·····m:···,·;mJ·······, mJ m WSDG -r-,..,...,...,r-r.,...,-r--r-r...,..~..,...rT..,...r"T...,.. ...,...,~ ..,...,...... o ($ VSOP - Model Pressure (rnb) Figure 9.5a (top) - Histogram showing number ofVSOP observations differing from model pressure ranging from -15 to 15 mb Figure 9.5b (bottom) - ... Numbers ofVSOP observations differing from i modelpressures by ship for pressure differences ranging from -20 to 25 °C. Shading indicates numbers of observations per 1mb range. 1·(l4() D. :0 rD -0.2 > ~(h4 '...... ---._...... o._...... _.-....- -....._...0100--'''''--____...... '94:0 Eignre9.6a{!qp) ·VSOP~measured 'pressurebinnedonmodeLpressme Figure9.6b{bottom) DjfferencesinVS-OP andOce'anWeatherShip Cumulus measure'd'pressures Irommodelvaluesbinnedonmodel pressure - 85- .....-.. .c ...... ,E CD ..:::s fI) Cl) CD a... "ii 0 :E" Q. 0 rn I > 9-80' 990 1000101-01020 1030 10-40 VSOPPressure-(mb) 1.0 0.5 O~O 1------4------...... ---1--...... ------.... -0115 (I) :3 0 :::s -1.0 - E o-:1 -1.5 .....u...w.a.~~ ~ .....-& .... 970 Figure 9.6 c {top) Differences in- VSOP measured pressures from-analysls and {orecast rnodelvalues .0 (mb)binnedopVSOPpressure o (rob). Figure 09.6 d(bottom) Differences in VSOP and Oce.an Weather Ship Cumulus measured pressures (mb)from an~ysisandforecastmodel v~uesbinnedoon modelpr~ssure{mb); - 86- 1..0 0.5 0.0 .. .. I. ~ ; • i i I Z -0.5 "-.-. I • I -1.0 I -1.5 I -2.0 -2.5 a... o Cl) > 5 10 15 20 Height of Temperature 1.0 0..5 :0g 0.0 ...CD :J -0.5 en (0 CD a- -10<0 D. -1.5 ..-0 0 -2.,0 :E -2.5 Q. 0 -3Jl en > -3.5 510 15 20 25 30 35 40 Height of Temperatur~ Sensor asl (rn) Figure 9..'1 a (top) - VSOP pressure difference frommodel analysis data (rob)' against heightoftemperature,sensor {m} by ship~ The line shows the expected. relationship uno pressure correction was applied Figure 9.1 b '(centre) - VSOP 'pressure diiIerencefrom modelforecast data (~) agairlSt height oftemperature sensor (m) by ship. Line as 9.7a Figure 9.1 c (bottom)- VSOP,pressure difference'from model analysis'data (rob) binned on difference in reference level from the ship·s mean reference level (m). The line shows the expected relationship if no correction is made for variations in ship·s draft