SMHI No 97, Jun 2001 Reports Meteorology and Climatology Kebnekaise -- Norwegian Sea Malung Klaipeda (%) 100 80 60 40 Viixjö 20 0 A N0AA AVHRR cloud climatology over Scandinavia covering the period 1991-2000 Karl-Göran Karlsson Cover: A1ea11 clo11d ji-eq11mcies i11 the rl}temoo11 i11 }116, (i11 rolo/11) f/11d the a11,11wl coursc ofdoud couerfor so,ne se!ectcd positions in the Scandi,Jtwir111 arcfl (rnrve plots) deriued ji-om fl te11-ymr NOM AVHRR rlfltf/ set 1991-2000. RMK No 97, Jun 2001 A NOAA AVHRR cloud climatology over Scandinavia covering the period 1991-2000 Karl-Göran Karlsson R eoort s ummary /Rapportsammanfattn1n2 lssuing .-\gerKy Utgi\·;m.: Report numher/Publikation Swedish Meteorological and Hydrological lnstitute RMK No. 97 S-601 76 NORRKÖPING Repor! date/Utgivningsdatum Sweden June 2001 Author (s) Förfallarc Karl-Göran Karlsson Title (rn1d Subtitle Titd A NOAA AVHRR cloud climatology over Scandinavia covering thc period 1991-2000 Abstract Sammandrag A tcn-year NOAA A VHRR cloud climatology with a horizontal resolution of four km has been compiled over the Scandinavian region based on results from near real-lime cloud classifications of the SMHl SCANDlA mode!. The frequency and geographic distribution ofthe cloud groups Low-, Medium- and High-level clouds, water and ice clouds and deep convective clouds have been studied in addition to the ten-ycar monthly means of total fractional cloud cover in the region. Furthennore, attempts to estimatc the diurnal cycle of cloudiness and typieal cloud patterns in various weather rcgimes (e.g., North Atlantie Oscillation phases) have been made. The cloud climate in the region was found ta be significantly affected by the distribution of land and sea. In partietdar. the Baltic Sea was shown ta suppress summertimc cloudiness substantially and this effect was shown to influence cloud conditions in major parts of the Scandinavian region. Huwcver, intcresting deviations from this cloudiness pattern were found 111 the Scandinavian mountain range, in the northern part af Scandinavia and over the Norwegian Sea. The quality af the sate! lite-bascd cloud information was examined by companng with corresponding surfa ce-o bserv a t ions given by SYNOP-based cloud climatologies for the same period. Results showed good agreemcnt but specific problems were found in winter. In addition, somc effects of the degradation of visible AVHRR channels were notieed. Comparisons have also been rnacle with internationally used global cloud climate data sets, namely the SYNOP-based CRU data set and the cloud climatologics from the ISCCP D2 series. K~·y words sök-. nyckdord Cloud elimatology, NOAA AVHRR. multispectral processing. cloud classification, validation of satellite-derived cloudiness Suppkrnentary 1wtes Tilliigg Numlwr ofpagec. ,\ntal sidor Language Spr,-ik 95 English ISSN and titk ISSN och titd 0347-2116 SMHI Reports Meteorology and Climatology Report .t\ailahle from Rapporten k.111 köpas fran: SMHI S-601 76 NORRKÖPING Sweden INTRODVCTION ....................................................................................................................................... l 2 THE SCANDIA CLOl:D CLASSIFICA TION MOD EL ......................................................................... 4 2.1 T!IEORIGINAL SCANDIA "IODEL- VERSION I ....... 4 2.2 THE "IODIFIED SCANDIA MODEL- VERSION 2 ...... 9 3 COMPILA TION OF SCANDIA CLOUD CLIMATOLOGIES ........................................................... 14 4 THE SATEL LITE DATA SET ................................................................................................................ 20 5 RESl:LTS FOR THE NOIWIC REGION - SCANDIA VERSION 1.................................................. 23 5.1 TOTAL CLOCD COVER AND CLOUD FREQUENCIFS ,s.,,, .... 23 5.1. I Seusom.d and JJwmh~r means m't'r the .?3 5.1.] The mmual (:rcle r!(ciouäiness 28 5. r3 lmer-wmuaf rariabilily <fclmuliness -· . 34 5. JA The diunud t:rdc u{cfoudiners .,,,. .37 5.2 TTIE SEPAR,\TION f\'.TO V:\RIOUS ClOLJD GROUPS ., ........... 40 5 . .?. I O1uu1m' c!oud groups ... ........ 4U 5_.?.:! St!mi-transparent Cirrus clowls. 41 5 . .:.3 h·c mu/ 1wt1er clmul'j .. .. 43 5.2.4 Fog and Stratus ... ........ 45 fl.5 Pr.:cij)ilating c/vuds and de-ep com·ec·!irt' clouds ........ 46 5.:!.n Fmctiona/ suh-pi.rci cloudiness ..... 411 6 RESULTS FOR THE :\'ORTHERN EUROPEAN REGION ·-SCANDIA VERSION 2.. ................... 49 7 \' ALIOATION RESULTS ......................................... ,.............................................................................. 52 7.1 CrndPARtso;,..;s WITI l S\'NOP OBSLR\"_-\TlONS ...... 52 7. I, J Vo/idatirm o/ cloml clinw10/ogiesJhm1 SCANDLJ J ·e1·sim1 1.. ... 52 ., ., 7. I.] hdidmion o{cloud dimu10/og)esJh)//1 SCANDJA Version :r ..... .. 6] l . .s.. COl\lP:\RISO;<s-\ \\TfH SOl.:\R R\f)L\TlON .\IE,.\SURE~H:NTS .............. .. ... 63 7.3 COMPARISONS \I 1TH IS(TP. CRU A~D Ui'L\ TE Sl\lt:LATION llATASETS. .. ... 67 8 THE IMPORTAi\'CE OF INHERENT CALIBRATION ERRORS IN THE NOAA AVHRR n.\TASET .................................................................................................................................................. 69 8.1 B,\CKGROL:'\D .\:',;D ST,\TLS OF TIil~ A 1/HRR C:\L!11Rt\ TION PROBLE/vl FOR \TSIBJ.F CIIANNFL':) ........ 69 R.2 IMPVT ON l llE SCANDIA CI.OUD CLI\I.\TOLOGY .... 70 9 DISCUSSfON ............................................................................................................................................. 74 9.1 Sll\•1)1.L\RY (;f :\i. flll~\"E1i-iF.'\ rs R[G\RDING SCANDIA CLOUD CLl\lATOLO(ilES .. .. ... 74 9) THE QL\LIT\ OF SCA:\iD!A CLOtlJ CLl\t,\TOI.OGIES ,. 75 9.3 FUTLRE PL\NS: THL Sl'(TESSOR TO SCJ\NDIA AND ENGAGFMl:;'\TS IN INTERNAflONAL Cll;\l:\TE ivfON[TORING PROGR·\\ll\lES ,.. .... 77 9.3. I Futun' usf:' uj'SCANDl:l dmul climatolugics .... 77 9.3 . .: The succf!ssor to SC'AND!A "'····· . ., .. 78 9.3.3 SJ!/-/[ t/ngagcmcm.,· in_filfun:: imemarimwl clmul c!imatc 111oniroring auh·itie.L.... ........ "'""·· 80 10 ACKNOWLEDGE\IENTS ...................................................................................................................... 80 Il ~Ff.'l'RE'<CES .......................................................................................................................................... 81 APPENDIX I. ACRONYM LIST ................................................................................................................... 84 APPENDIX 2. MONTHLY CLOUD CLl'1ATOLOGIES 1991,2001........................................................ 85 1 INTRODUCTION Measurements from high resolution imaging sensors onboard both geostationary and polar orbiting satellites have up lo the current <late been available lo the meteorological community for more than three decades. These sensors were particularly designed to enable monitoring of cloudiness and cloud systems and for monitoring of surface conditions over cloud-free areas. However, despite the produced wealth of information <luring this long period, only a limited number of studies on cloud conditions presenting results from long-term quantitative applications (i.e., climatologies) have been presented. This could be compared lo the relatively !arge number of studies concerning surface conditions (e.g., studies on surface parameters like NDVI - Normalised Difference Vegetation Index - and others, for example as reported by Gutman, 1989 and Glasser and Lulla, 2000). This is explained by the fäet that the rather short life cycle of clouds and cloud systems require utilisation of images with high temporal and spatial resolution as a contrast lo for example studies of surface parameters. Consequently, the required data amount from high-resolution imagers lo enable such studies 1s enormous. Quantitative efforts have so far been limited to the compilation of coarse resolution data sets on thc global scale lo be used e.g. in global climate studies. The most well known example here of such a satellite-based cloud climate data set is produced by ISCCP - the International Satellite Cloud Climatology Project (Rossow and Garder, 1993, and Rossow and Schiffer, 1999). This data set consists of a complete and consistent set of global cloud and radiance parameters derived from sensors on both geostationary and polar orbiting satellites. However, the focus on the global scale has forced the use of a quite limiting sub-sampling technique. Consequently, only a small fraction of the available global satellite data set has been utilised, both in tenns of the used spatial resolution and the number of used spectral bands of the sensors. The applied sampling strategy basically means that it is assumed that the dynamical evolution of cloud systems is randomly distributed within a larger segment. Consequently, it should be sufficient to select data from only a few high resolution pixels (for ISCCP with a spacing of about 30 km) within the segment and from this sub-set of data construct statistics from very long time-series of measurements lo describe mean conditions valid for the entire segment. This methodology is justified for areas with a weak dependence on local scale features (e.g., over oceanic areas) but for other areas (e.g., near coastlines or steep orography) the method is likely to give unrealistic results. Even if the interest on global climate studies is continuously high, an increasing attention on the cffects of climate change on the regional and local scale has been noticed lately. Many national meteorological services (NMS) and other agencies