Betting on Rain Managing Seasonal Risk in Western Nsw Betting on Rain

N S W D P I

BETTING ON RAIN MANAGING SEASONAL RISK IN WESTERN NSW BETTING ON RAIN

MANAGING SEASONAL RISK IN WESTERN NSW

R B HACKER Y ALEMSEGED P M CARBERRY R H BROWNE W J SMITH

RON HACKER, YOHANNES ALEMSEGED, PAUL CARBERRY, BOB BROWNE & WARREN SMITH OTHER TITLES IN THE RANGELAND SERIES published by Qld DPI

WEB ADDRESS FOR DOWNLOADING WHOLE OR PA R T O F betting on rain FROM THE NSW DPI WEB SITE www.dpi.nsw.gov.au/publications?a=63667 BETTING ON RAIN MANAGING SEASONAL RISK IN WESTERN NSW

R B HACKER, Y ALEMSEGED, P M CARBERRY, R H BROWNE, W J SMITH Title: Betting on Rain, Managing Seasonal Risk in Western NSW

Authors: R B Hacker, Y Alemseged, P M Carberry, R H Browne & W J Smith

Published by NSW Department of Primary Industries.

Copyright of this publication, and all the information it contains, jointly vests in the Land and Water Resources Research and Development Corporation, with its brand name being Land & Water Australia, and Australian Wool Innovation Limited, and the State of New South Wales (NSW Department of Primary Industries). These parties grant permission for the general use of any or all of this information provided due acknowledgement is given to its source.

First published May 2006

ISBN 0 7347 1727 X

This project was funded through Land, Water & Wool – a joint investment between the wool industry’s peak research and development body, Australian Wool Innovation Limited, and the nation’s premier investor in natural resource management research, Land & Water Australia.

Disclaimer

The information contained in this publication is based on knowledge and understanding at the time of writing (April 2006). However, because of advances in knowledge, users are reminded of the need to ensure that information upon which they rely is up to date and to check currency of the information with the appropriate officer of New South Wales Department of Primary Industries or the user’s independent adviser. job #4350 FOREWORD

Farmers and graziers in western NSW face special challenges in managing their businesses and their land for long-term viability. Not the least of these is posed by the extreme variability and non-seasonal distribution of rainfall. In such an uncertain environment decision making can be particularly difficult and the management of seasonal risk is a key to both business success and astute environmental stewardship. This guide will provide practical assistance with this important task. Betting on Rain draws together basic information on weather and climate systems, and recent research on the development and application of seasonal risk assessment tools, in the specific context of western NSW. The guide provides both informative insights into the factors that determine regional weather patterns and practical tools that will allow landholders to utilise what predictability does exist to improve management decision making. The research team worked with a network of over 300 landholders across western NSW. Their feedback has helped shape this booklet and I expect that they and others will find it a valuable addition to the earlier rangeland publications in the glove box guide series. Climate science can not yet provide answers to all the questions that landholders ask about future seasonal conditions. But it can make a practical contribution when the available tools are understood and properly applied. This publication will make a substantial contribution to that process. It should reward frequent reference whenever landholders grapple with the climatic uncertainties that confront primary production and land management in western NSW.

B D Buffier Director-General NSW Department of Primary Industries TABLE OF CONTENTS

INTRODUCTION 1 How to use this guide

PART 1 – WHAT DRIVES CLIMATE AND WEATHER? 3 Global circulation and weather 4 How do these pressure systems work? 6 What makes it rain? 10 Clouds 10 Reading ‘weather maps’ 11 Seasonal outlook indicators 13 What are El Niño and La Niña? 15 Some other influences on seasonal conditions

PART 2 – MANAGING CLIMATE RISK 19 How do seasonal risk assessments fit into management? 20 Some basic concepts 21 Probability-based assessments and the accuracy question 21 What level of probability will change a decision? 22 Using seasonal risk assessments in management 23 Taking a calculated risk

PART 3 – USING HISTORICAL RAINFALL DATA 25 Background 33 A comment on climate change

PART 4 – SEASONAL RISK ASSESSMENT 35 Background 36 SOI phases 40 Important note TABLE OF CONTENTS

PART 5 – TRIGGER POINTS 65 Background 67 Identifying trigger points for your property 70 Using trigger points and pasture growth profiles FOREWORD

PART 6 – SOURCES OF CLIMATE AND WEATHER INFORMATION 81 Short term weather forecasts 82 Weather forecasts on fax 83 Seasonal Climate Outlooks 83 Seasonal outlooks on fax 84 Accessing historical climate data

PART 7 – SOME PRODUCER’S VIEWS 85 Hugh McLean, Booligal 85 ed Fessey, Brewarrina 86 Peter Bevan, Broken Hill

ACKNOWLEDGEMENTS

GLOSSARY

APPENDIX 1 92 Variation over the year in the relationship between SOI phase and the probability of exceeding median pasture growth.

APPENDIX 2 93 Model versions (either the location for which the GRASP model was calibrated or an ‘average’ calibration) and the climate record used to produce the pasture growth profiles included in Part 5

FOREWORDINTRODUCTION

This guide describes the physical forces that drive weather and climate in western NSW, and tools that can be practically applied to help manage seasonal risk in this highly variable and uncertain environment. It presents the results of the project Improved seasonal forecasts for wool producers in western NSW, funded by Land and Water, Australia and Australian Wool innovation Limited through the Land, Water and Wool program, and NSW Department of Primary Industries (NSW DPI). As part of this project a team from NSW DPI worked in conjunction with colleagues from the Queensland Department of Natural Resources and Mines to analyse the value of seasonal risk assessment systems in western NSW, and with a network of some 330 producers to determine how the results might best be used in practical management. The research component evaluated the usefulness of seasonal risk assessments for both rainfall and pasture growth derived from the Southern Oscillation Index and Sea Surface Temperatures. It also developed the concept of ‘trigger points’ – key dates for management decisions – and identified these dates for a range of locations throughout the region. The interaction with the landholder network, both through project newsletters and face to face workshops, was invaluable in identifying those management decisions that are most dependent on seasonal outlooks, as well as the background information on climate and weather systems which would be of practical value to managers. That information has been included in this guide.

HOW TO USE THIS GUIDE This guide is intended to be used as a working manual. It contains seven major sections. While an initial reading of the whole book would be useful, practical application will generally require reference to specific sections. The colour coding and the description of each section below will assist this quick-look application.

Part 1 – What drives climate and weather? Basic drivers of weather and climate in western NSW; seasonal outlook indicators; El Niño and La Niña. (If you have previously attended a NSW Department of Primary Industries climate workshop you might skip this section although a refresher could be useful). Part 2 – Managing climate risk Some basic issues in managing climate risk using a probability or ‘odds’ based approach.

Part 3 – Using historical rainfall data Understanding and using long-term climate records; long-term climate data for locations in western NSW.

Part 4 – Seasonal risk assessment When are seasonal risk assessments useful?; maps showing the probability of exceeding median pasture growth in various SOI phases at these times.

Part 5 – Trigger points What are trigger points?; pasture growth profiles for various locations in western NSW.

Part 6 – Sources of climate and weather information Web sites and other sources of weather and climate information.

Part 7 – Some producers’ views Some producers’ opinions about seasonal risk management and use of the tools described in this book. INTRODUCTION PART 1 – WHAT DRIVES CLIMATE AND WEATHER?

• Global circulation and weather – high and low pressure belts, global wind patterns FOREWORD• What makes it rain? – sources of moisture and the processes that can trigger the formation of rain • Major weather systems that deliver moisture to eastern Australia or act as triggers – satellite images and synoptic charts, reading weather maps • Seasonal outlook indicators – the Southern Oscillation Index, sea surface temperatures, El Niño, La Niña and other factors – how these works to influence the seasonal potential Note – if you have already attended a NSW Department of Primary Industry climate workshop you might wish to skip this section.

Global circulation and weather Local weather is driven by the major global circulation systems. These in turn are driven by heat, provided almost exclusively by the sun. The sun provides most heat to the part of the earth which is closest, near the equator, and creates warm water and land in this strip around the globe (Figure 1.1). This, in turn, creates warm air which expands and rises resulting in a low pressure belt at the earth’s surface in the equatorial region. At altitude, this air cools and falls in a belt at about 30o S latitude, over Australia, resulting in a belt of high pressure at the surface. To forecast weather in Australia you need to watch the closest of the six to eight high pressure systems formed around the globe in this belt where the air is falling.

Figure 1.1 Atmospheric wind and pressure systems WHAT DRIVES CLIMATE AND WEATHER? PART 1 shown in the diagram. the in shown are patterns movement general The isobars. closer as up show will and speeds wind faster create will differences pressure Larger maps. synoptic on isobars of closeness the by seen is This speed. wind the determine systems the of closeness the and pressures the in difference of amount The 1.3b). and 1.3a (Figures high a in downward, and anticlockwise, and system pressure low a in upwards, and clockwise, spirals air the hemisphere work? southern the In systems pressure these do How winds of direction and strength indicate arrows of direction and size the – map circulation Wind 1.2 Figure equator. the winds’ west along to ‘trade east of region the and hemisphere southern the in circulations pressure high anticlockwise large five the Note forecasting. for conditions starting accurate ensure to recorded continually are measures Such day. specific one on was it as direction and speed wind actual the shows 1.2 Figure belts. pressure high and low the of position the in shift north–south a also and seasons, the produces sun the around earth the of rotation The hemisphere. southern the in winds trade the of direction south-easterly characteristic the and flow air the of deflection apparent the in resulting east, to west from earth the of rotation the by influenced also are patterns These patterns. wind global the creates that force basic the is out balance to try they as systems pressure the in air of Movement reached. never is equilibrium atmosphere, the in occurring cooling and heating continuous the With completed. is Cell Hadley the equilibrium, reach to attempt an in belt pressure low the towards flows belt pressure high the in air As

PART 1 WHAT DRIVES CLIMATE AND WEATHER? winds at ground level spiral inward and clockwise to lows winds at ground level spiral outward and anticlockwise from highs

spiraling, sinking air – clear spiraling, rising air – clouds Also, highs tend to be big, slow moving systems while lows are smaller and faster. Frosts and fog often form in the central regions of these slow highs in the colder months. determines the surface wind direction and thus whether it will blow over a good moisture source or not, before it gets to a trigger. Look to the highs for futureWhere weather. you are in relation to the centre of the high provides a guide to what is coming. The air going into lows can contain moisture and as it rises it cools and can form rain. Descending air pushing out from the highs is usually dry, the moisture condensed out when the air rose.While lows are often triggers for rain, it is the air pushing out from the highs that agricultural activity such as large fallow areas or irrigation. It is important to recognise that the air is moving up or down as well as round and either in or out, all at the same time. It is a very dynamic, three dimensional system. Air moving downward in a high spirals outward at an angle of about 15º to the lines of pressure, and air moving upwards into the lows spirals inwards at a similar angle.This will vary slightly with terrain and there will be local effects from hills, valleys, forests or It is most likely that air coming off the ocean, particularly the warmer parts of the ocean, will bring moisture. Air from central Australia (no free water, few plants) will be dry. Air coming from the Antarctic will be cold. Air from the north will be warm. Figure 1.3a (left) Directions of air flow, Figure 1.3b (right) A side view of how highs and lows behave The location of the pressure systems can give a good guide to what will come with the wind. WHAT DRIVES CLIMATE AND WEATHER? PART 1 • • • of because rises Air elevation. in m 100 each for 0.65ºC about of cooling in resulting rise, to air cause that things are triggers common most The moisture. of plenty contains air the if even occur will rain no trigger a Without rain. for trigger the as act can cool to air causes that Anything humidity. relative the on depends point dew the at temperature The grow. and form droplets the water at ‘dewand point’over cross they until rate, condensation the than faster decreases rate evaporation the drops temperature the As cool. to needs air moist that is important most occur. to The this for met be also must conditions other of number A rain. produce to sufficient not is moisture of source good a having Unfortunately, onshore. comes system the where to closest are which regions in time to time from falls good produce still systems these although precipitation, for available moisture less have will Ocean Southern the from originating air Cooler moisture. of source major the provides waters ocean warmer the from evaporation rule general a As trigger. a and moisture of source a rain: it make to required are Twofactors rain? it makes What Figure 1.4 An example of warm air uplift by a cold front cold a by uplift air warm of example An 1.4 Figure Figure 1.4 below. These generally produce widespread rain. widespread produce generally below. These 1.4 Figure in shown as air warmer undercutting fronts cold e.g. systems pressure air of Collision rain. not would otherwise that systems from showers in results often mechanism This cooling. some cause to enough up pushed get will air some turbulence, enough create which things are there If turbulence. frictional or mechanical – earth’ssurface the on Irregularities rising. without them around flow simply not can air the so enough wide and cooling significant cause to enough high be must hills The hills. the over get to rise must air The lift. Orographic – mountains and Hills COLD AIR COLD WARM AIR WARM PART 1 WHAT DRIVES CLIMATE AND WEATHER?

0000 UTC 08 JUN 2002 10AM EST 08 JUN 2002 Sunset, where the source of heating is removed allowing the air to cool. A trough of lower pressure where the air expands as it flows out of the region of high pressure to a zone where pressure is less, spreading the heat it contains and thus cooling slightly.This can be just the region between two highs but a more significant trough with strong air flow is usually necessary to produce substantial rain. Figure 1.5 shows a modest trough (dashed line) through the north east of NSW. Mixing with cooler air as two circulating systems interact. Both systems need to be moist to produce rain.This combination can result from the interaction of a slow moving high in the Bight and Tasman high a but as a trigger this mechanism is less effective than an undercutting cold front pushed in from the south west. Hot air uplift – local convection, resulting in storm rain in summer. Sufficient heat is required to create a local hot spot which can produce a pool of warm, rising Mostair. storms occur in the afternoon, following the hottest part of the day. A circulating low – convergence pushing air up. As explained earlier, the air in a low pressure system is rising, with air converging into it from surrounding systems. If the air is moist, the rise within the low can trigger rain.

MSL Analysis NATIONAL METEOROLOGICAL & OCEANOGRAPHIC CENTRE BUREAU OF METEOROLOGY VALID

Figure 1.5 Synoptic chart showing trough through north east NSW • • • Other causes of cooling can be • • WHAT DRIVES CLIMATE AND WEATHER? PART 1 sometimes bring extensive rain to inland, southern and eastern Australia. Australia. eastern and southern inland, to rain extensive bring sometimes They rain. of chances the increase will front undercutting an as such features other with Interaction rainfall. widespread produce itself, in may, that mechanism a is ascent scale broad This 1.7). (Figure rises and southeast moves Australia, of Western coast the off forms band The Ocean. Indian northern the in originate they source, winter / autumn an Usually 2. Agency. Meteorological Japan the of (GMS-5) Satellite Meteorological Geostationary the from Meteorology of Bureau the by processed originally image Satellite Source: thunderstorms. into develop can vapour This pictures. the on show not does which vapour become has it when even NSW, western to over drifts often moisture down, break systems these As Australia. central and west north over coast west north the off cyclone the from cloud with pattern cyclone summer typical A Figure1.6 rains. general more produce to front a with interact will moisture this occasions rare On fire. cause can which lightning have may but rain no or little very produce storms Some areas. small quite in hail or wind strong sometimes and rain patchy in result often These storms. as typically 1.6), (Figure elsewhere rain produce to triggers other with interact can and down breaks cyclone or monsoon the after remains some but system the within rain as falls Some atmosphere. the entering moisture of volumes huge in result temperatures warm where seas tropical the in originate phenomena summer These 1. triggers. as act or Australia eastern to moisture deliver systems weather major Several

VALID MSL Analysis MSL METEOROLOGY OF BUREAU CENTRE OCEANOGRAPHIC & METEOROLOGICAL NATIONAL

North west cloud bands from the northern Indian ocean – April to August to April – ocean Indian northern the from bands cloud west North April to November – regions tropical the in cyclones and Monsoons 10PM EST 14 DEC 1999 DEC 14 EST 10PM 1200 UTC 14 DEC 1999 DEC 14 UTC 1200

PART 1 WHAT DRIVES CLIMATE AND WEATHER?

S-9) of the Japan Meteorological Agency. E

0000 UTC 19 JUN 2005 0000 UTC 19 OCT 2003 10PM EST 19 JUN 2005 10AM EST 19 OCT 2003 The Tasman Tasman high/coldThe front system – May to October

NATIONAL METEOROLOGICAL & OCEANOGRAPHIC CENTRE BUREAU OF METEOROLOGY MSL Analysis (hPo) MSL Analysis NATIONAL METEOROLOGICAL & OCEANOGRAPHIC CENTRE BUREAU OF METEOROLOGY VALID VALID

Geostationary Meteorological Satellite (GO Source: Satellite image originally processed by the Bureau of Meteorology from the Geostationary Meteorological Satellite (GMS-5) of the Japan Meteorological Agency. Figure 1.8 Cloud pattern and synoptic chart showing a typical Tasman high/cold front system at work the rain. Although this rain moves through from the west, the moisture has actually originated in Tasman theand Coral seas and has been brought over the land mass by the easterly and north-easterly air streams produced by the high pressure system. Rainfall is often widespread through amounts may vary. The fronts themselves are poor sources of moisture but the uplift they create when they interact with the warmer air circulating from Tasman thehigh is the trigger which produces frequently interact with cold fronts over southern Australia to produce widespread ‘frontal’ rain.The cold fronts circulate continuously but it is only in winter, when the global circulation pattern moves to the north, that they intersect the Australian land mass. How far they move to the north is an important determinant of winter rainfall in northern NSW. 3. High pressure systems in Tasman theSea (Figure 1.8) bring moist easterlies onto south eastern Australia throughout the However, year. from May to October these systems Source: Satellite image originally processed by the Bureau of Meteorology from the Figure 1.7 A strong north west cloud band streaming in from near Indonesia WHAT DRIVES CLIMATE AND WEATHER? PART 1 10 This usually means a good depth of moisture at a concentration which is already producing producing already is which concentration a at moisture of depth good a means usually This clouds. high of tops the usually are white) as (shown parts cold the region, Australian the of image infrared an just is it While moisture. the to clues the The provides ‘cloud’picture region. your in is it while it with interact to mechanism trigger some for and days few next the in way your come could which atmosphere the in moisture of indicators for Look region. your in active potentially are triggers what and is moisture the where out find to used be can services) fax in or on TV newspapers, in shown (as maps weather and pictures Cloud maps’ ‘weather Reading rain. into vapour more turn to force, in remain to condensation inducing conditions the and time, some for continue to air moist of inflow the requires agriculture to useful amounts in rainfall Toground. produce the on deep mm 1 be only would high kilometre 1 cloud a in water visible the all that means This vapour. uncondensed is which portion larger a is there though ice, or water condensed is cloud a of volume the of million per part 1 about Only place. taking is activity condensation that sign visible a just are They them. from falls which water the all contain not do Clouds Clouds (GO Satellite Meteorological Geostationary the from Meteorology of Bureau the by processed originally image Satellite Source: coast. NSW southern the off a low Tasman into drawing Cloud 1.9 Figure 1.9). (Figure areas inland affect not do but winds strong and rain intense bring can lows These winds. southerly cold with association in often east, south the in and coast mid the on only rainfall affect systems These 4. fronts and strong trough zones are good triggers for inland regions. regions. inland for triggers good are zones trough strong and fronts Cold triggers. any to clues the provides it) on drawn lines isobar the (with chart synoptic The rain. of source potential a – condensation

VALID MSL Analysis (hPo) Analysis MSL METEOROLOGY OF BUREAU CENTRE OCEANOGRAPHIC & METEOROLOGICAL NATIONAL

Tasman lows – any time but mostly June to September to June mostly but time any – lows Tasman 10AM EST 10 JUL 2005 JUL 10 EST 10AM 0000 UTC 10 JUL 2005 JUL 10 UTC 0000

E S-9) of the Japan Meteorological Agency. Meteorological Japan the of S-9)

11 PART 1 WHAT DRIVES CLIMATE AND WEATHER? southern oscillation index oscillation southern 0 30 20 10 -20 -10 40

2000 1990 1980 1970 1960 1950 1940 1930 – of specific things occurring, thus providing a good basis for l Niño Southern Oscillation phenomenon: Annual average rainfall for the 1920 E probability 1910 1900

900 400 300 200 700 600 500 800 annual average rainfall average annual (orange line). Source: AustralianWeather Calendar 2004, Commonwealth Bureau of Meteorology and Australian Meteorological and Oceanographic Society. Figure 1.10 Studying the Murray-Darling Basin (solid grey) is highly correlated with values of the Southern Oscillation Index The SOI is just an indicator and does not include all the factors influencing rain in eastern Australia so while the general relationship between SOI and rainfall is useful – as can be seen from Figure 1.10 – there is considerable variation that is not simply determined by the SOI. Australia. In years with a positive SOI rainfall is commonly above average.When the SOI is negative the trade winds are usually weakened and the rainfall in eastern Australia can often be below average. indication of what is happening with the atmospheric circulation across the Pacific. When the SOI is positive the trade winds typically blow strongly across the warm western Pacific Ocean and pick up plenty of moisture which can then lead to rain over eastern The SOI is based on measures of the difference in air pressure between Darwin Tahiti. and The difference is compared to the long-term mean difference at the same time of year and is expressed as a number which ranges from about -30 to +30. It is a method of getting an usually specific enough to predict exactly what will happen.They are used to get a guide to the chance – or risk assessment but not a definitive prediction. One is the Southern Oscillation Index or SOI. A variety of indicators can be used to look a bit further out than just the next week or so, which is all that can be done with weather forecasting systems at the moment. These indicators are known to be connected to major shifts in weather patterns but are not Seasonal outlook indicators 1 WHAT DRIVES CLIMATE AND WEATHER? PART 1 2 and one when they were low. These are known, respectively, as the the as respectively, known, are low. were These they when one and high were Pacific eastern and central the SSTsin the when period one shows 1.12 Figure rain. the produce that systems atmospheric interacting the and supply moisture potential the both to guide a be can oceans the of state the season, any over rain trigger may systems atmospheric different of number a While oceans. the in changes by modified are themselves systems atmospheric the addition, In temperature. water on dependent largely is supply this and rain create systems atmospheric the which from moisture the supply oceans The data. worldwide collect to place in put were satellites when 1982 since measured been have temperatures is indicator outlook seasonal Another elsewhere. from different region your in rainfall and SOI between correlation the make may effects local Furthermore, probabilities. rainfall different with associated pattern, different vastly a into settle and future immediate the in jump may it as autumn or summer late the in used be not should It jump. a after settled has it once months few following the in rain for potential the to guide useful a be can it that mean low, is SOI the when lower and high is SOI the when higher be to rain for tendency the to added when features, two These SOI. the of average moving 30-day the of values end-of-month the shows graph The 2004. to 1996 from SOI 1.11 Figure 1. SOI: the of behaviour the of features important two shows 1.11 Figure conditions, which are explained further in the next section. next the in further explained are which conditions, 2.

SOUTHERN OSCILLATION INDEX the SOI tends to jump each autumn and settle into one of three patterns – high, low or or low high, – patterns three of one into settle and autumn each jump to tends SOI the once it has jumped it tends to remain in the same pattern for many months. months. many for pattern same the in remain to tends it jumped has it once middle; the about wobbling –40 –30 –20 –10 10 20 30 0 JAN 96 JAN CO MM ON JAN 97 JAN W EALTH B UREAU JAN 98 JAN

OF M ETEOROLOG JAN 99 JAN Y JAN 00 JAN Sea Surface Temperature Surface Sea 5 M JAN 01 JAN M ONTHL ONTH Y Y M SOI EAN JAN 02 JAN (W EIGHTED or SST. Sea surface surface SST. Sea or ) E JAN 03 JAN l Niño and La Niña Niña La and Niño l

JAN 04 JAN 13 PART 1 WHAT DRIVES CLIMATE AND WEATHER? A

2 > A

2 – 5 1. A

5 1–1. A –1 5 0. A

5 0–0. –0 B 5 0. B 5 1–0. l Niño, meaning‘boy asthechild’ the event occurred near Christmas time. E –1 B 5 1. B 5 –1. 2 B 2 < ach year around Christmas time the waters off the Peruvian coast warm a little.The locals l Niño conditions (September 1982) La Niña conditions (September 1988) in a corresponding air pressure gradient at the surface, with higher pressure in the east over in the east over the surface, at with higher pressure gradient air pressure in a corresponding dry waters the eastern cool, The air above in the west. pressure and lower water the colder rainfall or drought in Australia, particularly eastern Australia. are Pacific equatorial and eastern in the central SSTs Normally, like this. works process The results gradient temperature This the north to end, east of Australia. the western than at colder produced flooding rain and poor anchovy catches (Anchovy prefer cooler water).They called these events These same conditions that bring heavy rain to the Americas are associated with reduced What are El Niño and La Niña?E noticed that at intervals of around 3 to 6 years the water became unusually warm.This such outlooks they are not absolute predictions but are given in terms of probability – the percentage chance of getting rainfall above the median. likely moisture supply for several months ahead. ‘Seasonal climate fromoutlooks’ the Bureau of Meteorology are based on the relationship between the Australian climate and the oceans, particularly ocean temperatures. As with all to come our relativeway, to this recent past. In contrast to atmospheric indicators, changeSSTs fairly slowly, because the ocean is a huge mass of water with high thermal density.They can therefore provide an indication of the derived from data between 1982 and 2004.This is not a long time in geological history and we may not have seen all the natural variation yet but the images do give us a look at the current SST in relation to the recent past and therefore a guide to the potential for moisture In these images the yellow to red areas indicate above-average temperature, the blue tones are areas of lower than average temperature, remembering that the average can only be Figure 1.12Variation of sea-surface temperature from average Source: www.LongPaddock.qld.gov.au B – below average, A – above average E 1 WHAT DRIVES CLIMATE AND WEATHER? PART 1 4 during such an event is shown in Figure 1.14. Figure in shown is event an such during pattern circulation typical The Australia. eastern in rainfall average above with associated be can and Niña La a called is event This Pacific. eastern and central the in occur can water cold of accumulation abnormal an and strong extremely become can winds trade easterly the an After and Variability Climate Source: an during pattern Circulation 1.13 Figure an during pattern circulation typical The Australia. eastern in especially average, than less often is rainfall and highs, the Tasman by particularly Australia, over systems weather into fed moisture of amount the in reduction a is result The Pacific. western the in up pick can they moisture of amount the reduces also temperature reduced the while winds, trade the of flow the reverses even or reduces gradient pressure the in change associated The fall. Pacific western the in those and rise, an During over theAustralian landmass(e.g. ofa by theaction Tasman high)itcanproduce rain. winds, whichpickupmoisture astheymove over thewarmerwater. thismoisture If isbrought thus flows toward westward alongthesurface thewarmerwestern Pacific, creating thetrade H E l Niño event, weather conditions usually return to normal. However, in some years years some in However, normal. to return usually conditions weather event, Niño l E l Niño, sea surface temperatures in the central and eastern equatorial Pacific Pacific equatorial eastern and central the in temperatures surface sea Niño, l DARWIN E l Niňo, Bureau of Meteorology 1994, Commonwealth of Australia. of Commonwealth 1994, Meteorology of Bureau Niňo, l E l Niño event Niño l E l Niño is shown in Figure 1.13. Figure in shown is Niño l PACIFIC OCEAN PACIFIC TAHITI H 5

1 PART 1 WHAT DRIVES CLIMATE AND WEATHER? l Niño but the E H TAHITI l Niño and La Niña is certainly not restricted E PACIFIC PACIFIC OCEAN TRADEWINDS l Niňo, Bureau of Meteorology 1994, Commonwealth of Australia. E DARWIN qually, many but not all periods of high rainfall have been associated E H to moisture supply as it traverses the tropics. Unfortunately the phenomenon has very little identified influence in western NSW as it rarely feeds systems which are the most common sources of rain in these regions. The location of the MJO can be linked with rainfall patterns, particularly in north eastern Australia. Although other synoptic factors ultimately determine rainfall, the relative position of the MJO can help forecast increased or decreased chance of rain, through its contribution It does not show up as a clear low in synoptic charts and can only be detected with sophisticated measuring and calculating techniques but it appears to assist the lifting of tropical moisture to where it can feed into Australian weather systems. The MJO (also commonly known as the 40-day wave) is a tropical atmospheric phenomenon first recognised in the early 1970s. It is a small region of comparatively low pressure which develops over the Indian Ocean and travels east across the tropics with a timescale ranging from 30 to 60 days, and a frequency of 6–12 events per year. majority have been. The Madden-Julian Oscillation (MJO) to northern Australia as some seem to believe. Some other influences on seasonal conditions with La Niña. The major impact of these events can be seen from Figure 1.15a and b. Note that the impact on western NSW can be high – the influence of Not all major droughts in eastern Australia have been associated with an Figure 1.14 Circulation pattern during a La Niña event Source: ClimateVariability and 1 WHAT DRIVES CLIMATE AND WEATHER? PART 1 6 Figure 1.15a Impact of major major of Impact 1.15a Figure WET 1941 1905 1994 1982 AVERAGE E l Niño events on Australian rainfall Australian on events Niño l 1997 1987 1965 1914 DRY CO MM ON W EALTH B UREAU 1940 2002 1991 1977

OF M ETEOROLOG Y 7

1 PART 1 WHAT DRIVES CLIMATE AND WEATHER? Y ETEOROLOG M OF

1955 1973 1988 1917 UREAU B EALTH W ON MM CO DRY 1975 1971 1916 1950 AVERAGE 1974 1956 1938 1910 WET Figure 1.15b Impact of major La Niña events on Australian rainfall 1 WHAT DRIVES CLIMATE AND WEATHER? PART 1 8 of its influence was possible in the preparation of this guide. this of preparation the in possible was influence its of account no but assessments risk seasonal improved to lead eventually may understanding complete more A forecast. be to fluctuations its allow not does it demonstrated, be to influence its allows phenomenon this of understanding current the while Unfortunately, between relationships different have regions other However, drought. severe particularly produce to seems IPO of coincidence the NSW west ‘warm’ or north ‘cool’In phase. a in is IPO the whether on depending modified be may rainfall Australian on phenomena the of effects the that are Indications scales. time decadal roughly on changes temperature surface sea to related is phenomenon The Zealand. New and Australia in interest research much of subject the currently is and 1999 (IPO) in named first was IPO The Oscillation Pacific Interdecadal

E l Niño and the IPO so the overall situation is complex. is situation overall the so IPO the and Niño l E l Niño with a cool phase of the the of phase cool a with Niño l E l Niño and La Niña Niña La and Niño l PART 2 – MANAGING CLIMATE RISK

• How do seasonal risk assessments fit into management? – tactical stocking decision or in-crop management decisions will benefit most FOREWORD • Some basic concepts for interpreting weather records and understanding climate risk assessments – mean, median, deciles and probability • Probability-based assessments and the accuracy question – probability-based assessments are neither right nor wrong, they just provide the odds • What level of probability will change a decision? – many people would like at least 70% but smaller shifts could still be useful • Using seasonal risk assessments in management – they’re only one part of the jig saw but they can help make progressive adjustments • Taking a calculated risk – even relatively small shifts in the odds may make a difference; its worth doing a simple calculation to find out

As part of the project we conducted surveys of wool producers in the Western Division and adjacent parts of the Central Division to determine how seasonal risk assessments might be incorporated into practical management. We also obtained feedback from our co-operators in response to newsletters, and from workshops. Some of the key issues raised by this interaction are discussed in this section.

How do seasonal risk assessments fit into management? For livestock enterprises we found that the annual cycle of husbandry operations (e.g. time of joining, shearing etc.) was not likely to be influenced by assessments of medium term (3–6 months) climatic conditions. These dates are determined by other major factors in the production system such as the need to have lambs at a certain age by shearing, or the need to have sheep in short wool when grass seeds are likely to be a problem. Understandably, these major strategic decisions are not influenced by medium term seasonal outlooks. However, tactical decisions – particularly those related to livestock sales and other drought management decisions, and livestock purchases – are potentially very open to influence by seasonal outlook assessments. For mixed farmers, major cropping decisions at the start of the season (e.g. the area to be planted) were found to be more influenced by actual conditions such as the depth of soil moisture or the quality of the seasonal break than by medium term climate outlooks. However, in-crop management decisions e.g. whether to apply additional nitrogen or to graze off a crop headed for failure could be influenced by seasonal outlooks. 2 MANAGING CLIMATE RISK PART 2 0 100 % (absolute certainty of success) or 0% (absolute certainty of failure) there is always a always is there failure) of certainty (absolute 0% or success) of certainty (absolute % 100 is probability the unless for uncertainty of idea the embodies inevitably – occurs event the before meaning has only which – probability of concept The 30%. is range decile 3rd the of top the than less or to equal be will number selected randomly any that probability the Likewise 50%. is median the than greater be will number selected randomly any that probability the numbers, of set any In observations. of number total the of percentage a as observed is event particular a times of number the expressing by calculated is and 0–100%, from percentage a as expressed usually occurring, event some of chance the simply of concept The separately). done be to has analysis each – value yearly the get to range decile any in values monthly the all up add simply can’t you that (Note median. the is range decile 5th the of top the represents that value on. The so and 2, range decile in are cent per 10 next the 1, range decile in are cent per 10 lowest The highest. to lowest from order in ranked are month) or year (by received rainfalls the All record. historical Deciles records. of years 50 least at have to desirable is it figures the in Toconfidence figures. have misleading produce can years few A important. very is based is median or average the which on years of number The months. individual for greater much be can difference the and average the than less cent per 10 about is median the Australia, in locations most at rainfall annual For below. cent per 50 and it above years of percent 50 the use to is rainfall annual describing of way better A average. misleading a in result can climate dry a in rain flooding of year One rainfall. annual the for true is same The average. the distort can income high very a with individual one that is reason know. The you people many of wage the exceeds wage average the that wage. Younotice may average the is example common A misleading. be can average the numbers of sets some or with However, ‘expected’value. the ‘typical’ describe to trying when readily most of think to tend we number the It’s set. The include These understood. be to need terms basic these of few A management. risk about discussion any in up crop inevitably terms statistical and concept statistical or mathematical a is Risk concepts basic Some decisions these at directed particularly are booklet this of remainder the in discussed are that tools The assessments. risk seasonal term medium from benefit to likely most those are decisions tactical then, operations livestock and crop both For would be at best 4:1. best at be would a for ‘win’ bookmaker a from get would you odds The 1:4. are success of odds the occur, not will it that chance cent per 80 an therefore and occurring, event some of chance 20% a is there example, for If, are. they what odds’, essentially is as which ‘bookmaker’s probabilities of think to useful be may it fact, In win. will favourite odds-on an that guaranteeing of way no is there as just trial, any of result the be may failure or success either that chance average give an even better idea of how dry or wet a month, or year, has been relative to the the to relative been year,has or month, a wet or dry how of idea better even an give average is simply the sum of a set of figures divided by the number of figures in the the in figures of number the by divided figures of set a of sum the simply is probability (or mean), mean), (or is fundamental to any discussion of risk management. This is is This management. risk of discussion any to fundamental is median , , decile and and probability median . . This is the value that has has that value the is . This 1

2 PART 2 MANAGING CLIMATE RISK

in either the top or bottom 20 per cent of the historical record) is about 70% to 80% before they would change tack from normal management. for a more extreme whatyear, level of probability would be necessary before you changed your Themanagement?’ response shown in Figure 2.1 indicates that the majority of respondents would like to know that the probability of an extreme year (in this case a year What level of probability willDuring the project co-operatorschange were‘if youasked could get a forecast showing probabilitiesa decision? the result you want. Nevertheless, making a decision on the basis of known probabilities is a lot better than just flipping a coin – unless you know that the probability really is about 50 per cent. Remember, failure of the favourite to win, or the failure of good seasons to eventuate despite a high probability, does not mean – even‘inaccurate’ that theor ‘wrong’ odds were with a 90% chance of exceeding the median there is still a 10% chance that you won’t get However, if the probability is above 70%, or less than 30%, we can feel a lot more confident in making a decision than we could be if we just flipped the coin. It is under these circumstances that probability-based forecasts really have something to offer. SOI has very little influence on pasture growth for the time and place we are considering. mightYou as well flip a coin or better still look at the long-term rainfall records (see Part 3) to work out the probability of certain amounts of rain. For For example, if the probability of exceeding median pasture growth in the next three months is around 50% (or in practice between 40 and 60%), you could not very confidently take a decision that depended on getting above-median growth even though the probability is correct.The historical record is simply telling us that the particular state of the about ‘accuracy’ aboutbut the level of probability necessary for you to feel confident in making or changing a management decision. on history, of a particular result occurring. During the project, many producers expressed concern about‘accuracy’ of the the forecasts or risk assessments. But the discussion above makes it clear that the question is not really However, at any specific time there is no way of guaranteeing that a particular result will materialise even if the probability is high,These risk assessments are not predictions that can beThey assessed simply with provide‘wrong’. hindsight the or‘right’ chances, as based rainfall or median pasture growth for that period.When these probabilities are derived from historical records i.e. actual rainfall records or the modelled pasture growth profiles derived from historical rainfall and other environmental indata, the ‘accurate’ theysense are that they have been correctly calculated. The seasonal risk assessments discussed in Part‘probability-based 4 are assessments’. They simply provide the probability of certain seasonal conditions being experienced in a given time period. In our case, they are expressed as the probability of exceeding median Probability-based assessments and the accuracy question MANAGING CLIMATE RISK PART 2 22 These managers also accept that sometimes their decisions will turn out to be wrong – that that – wrong be to out turn will decisions their sometimes that accept also managers These along. go they as outlooks updated getting and up shows it when stress the reduce to adjustments further making stressed, being are things when of warning early get to system production the of monitoring the increasing by and early management to adjustments small making by outlooks of sorts these use industries other in managers Nevertheless, occur. will want don’t we result the that chance (known) a always is there since outlooks) climatic range medium for possible were these (assuming predictions using than difficult more much is probabilities using Furthermore, decisions. management most determines that data of jigsaw the in information of piece one management only are assessments in risk seasonal Probability-based assessments risk seasonal Using 2:1. of odds has still level this as 65%) (about probabilities lower at implemented be may that changes small for look managers that suggest Wewould wise. very be would health plant and moisture soil stock, of monitoring intense and changes management early some do, they When often. very occur to going not are magnitude this of probabilities (P. Leith, probabilities lower on act to prepared are assessments risk seasonal with familiar are who those that suggests evidence recent some although changed, be would management before 70–80% least at of probabilities see to like would producers of majority the that therefore Weexpect percentiles. 80th or 20th the not percentile), 50th the (i.e. value median the exceeding of probability the on based are booklet this in provided assessments risk The decisions management 2.1 Figure production system than you will pass up in lost opportunities. lost in up pass will you than system production the repair or up prop to having not by more much save will you and wrong are you than often more right be will you outlooks the to according things adjust you if that know they because this accept to ready are They profit. and production extra for opportunity an missed probably have they that and necessary not really were changes management early the

% respondents E personal commumication personal 10 15 20 25 30 ffect of seasonal risk probability on the proportion of wool growers willing to change change to willing growers wool of proportion the on probability risk seasonal of ffect 0 5 change decision @ decision change 55 1 60 0

… 65 2 , 2004) . Given the current risk assessment capability, capability, assessment risk current the Given . 2004) , 13 70 1 75 4 1 80 6 85 6 1 nu mb 90 3 er of respondents of er 95 2 pro b a b ilit 100 1 y 3

2 PART 2 MANAGING CLIMATE RISK

is is carrying on changing (20,000 x 0.65) + (5000 x 0.35) = $14,750. The ‘expected benefit’The of (20,000 x 0.35) + (10,000 x 0.65) = $13,500. ‘expected benefit’The of Y ILIT B A 0.35 0.65 0.65 0.35 B PRO $ E M 5,000 INCO

20,000 10,000 20,000 NET Carry on, correct decision Carry on, incorrect decision Change, correct decision Change, incorrect decision OPTION normal ($10,000 compared with $15,000) when you consider the chances of the various outcomes the least risky decision is still to make the change. So even though it would be tempting to take the risk on a smaller cost by carrying on as the right decision. However, if you sell early and it turns out to be the wrong decision then you will incur a cost of $15,000 and end up with a net income of only $5000. So the possibilities you are facing are: poor reproduction from retained stock so the net income is only $10,000. Lets assume, for the sake of simplicity, that if you decide to sell early and it turns out to be the right decision your net income will be same as if you carried on normally and that was If you carry on normally and the season turns out OK you expect a net income of $20,000 from sales but if the season turns out dry it will cost you $10,000 in feed, lower growth and Using a livestock example, say the SOI Phase is indicating a 65% chance that the next three months will be dry and you look at the choice between carrying on normally or selling half your usual sale stock early. Almost everyone uses this phrase but few people have actually taken a calculated risk. do soTo you need to estimate the likely net returns from each of the possible outcomes and the odds of each of those outcomes. reduce grazing pressure or finding agistment for your core stock. Taking a calculated risk necessary, purchasing and storing some fodder while it is relatively cheap, talking to your financier about the possibility of accessing more funds or rescheduling payments, destocking small areas to allow pasture or browse to freshen up, harvesting feral animals to but important changes could sensibly be made at slightly lower probabilitiesThese may include things like selling a portion of normal annual sales at earlier times, changing shearing date, if possible, to have stock in a condition which will allow sales if this becomes if reasonable, make a small adjustment to your plans. Monitor both the paddock and the updated outlooks and revise the plans regularly. While it would be nice to always have probabilities of at least 70–80 per cent, many small Agricultural managers can use the same techniques. needYou to carefully assess the actual situation in the paddock, add in the chance of the coming season being wet or dry and, MANAGING CLIMATE RISK PART 2 24 side of the alternative actions and compare them. The result might surprise you. surprise might result The them. compare and actions alternative the of side each on risk, or benefit, expected the estimate to need you risk calculated truly a To take (change). $16,250 and normally) on (carry $12,500 be would example above the in benefits relative the then 75% is season dry a of probability the if example For rapidly, increases change the of benefit the so increases, odds in shift this As costs. in difference the outweigh to enough big is odds in shift This 2:1. about to evens from is This 65:35. to 50:50 from is change the fact in but great very seem doesn’t 65% to 50% of base assumed an from odds in shift The

PART 3 – USING HISTORICAL RAINFALL DATA

• Analysis of historical rainfall data is particularly useful when other seasonal indicators are unreliable FOREWORD • For the moment historical data still provide the best means of describing the climatic characteristics of a location despite the evidence for global climate change • Diagrams and tables summarising historical rainfall data for 12 locations in western NSW

Background Historical rainfall data can provide a useful guide to the range and variability of weather in the medium-term future; especially at those times when the seasonal assessments discussed in Part 4 are unreliable. A history of 100 years or more will probably encompass almost the entire range that will be experienced over the next 10 years. While we need to be mindful of the possible effects of climate change (see below) a record of this length provides the best indication we have of the likely frequency and magnitude of extreme events, though it can not provide any real clues about when they might occur. Monthly median values provide a good description of the long-term climate at any location and provide a reference point for seasonal forecasts that are expressed in terms of probability of exceeding the median (though remember that the median for a period of several months is not the same as the sum of the monthly medians). Monthly medians are summarised for 12 locations in western NSW in Figure 3.1. Monthly deciles are particularly useful as they allow you to estimate the probability – the ‘bookmaker’s odds’ – of receiving any amount of rain in any month of the year. If you feel inclined you might even adjust the odds slightly according to the SOI at times of the year when it is a good indicator (as discussed in Part 4). The tables in this section provide monthly rainfall deciles, and other information, for the 12 locations shown in Figure 3.1. The decile values are presented as the % of years in which rainfall has been equal to or greater than the figure shown. At Bourke, for example, January rainfall has been equal to or greater than 29 mm in 40% of years. In other words there is a 40 per cent chance of getting 29 mm or more of rainfall in January, based on the historical record. The standard deviation given in the tables is a measure of the variability of rainfall from year to year – the higher the figure the more variable the rainfall. The standard deviation expressed as a percentage of the mean (called the Coefficient of Variation) is a good way of comparing the relative variability from month to month. USING HISTORICAL RAINFALL DATA PART 3 26

MEDIAN RAINFALL (mm) MEDIAN RAINFALL (mm) MEDIAN RAINFALL (mm) MEDIAN RAINFALL (mm) Figure 3.1 Median monthly rainfall for locations in western NSW western in locations for rainfall monthly Median 3.1 Figure 10 20 30 10 20 30 10 20 30 10 20 30 TIBOOBURRA WENTWORTH HILL BROKEN WALGETT BROKEN HILL BROKEN TIBOOBURRA F A J A O D N O S A J J M A M F J WENTWORTH WILCANNIA IVANHOE BOURKE HAY HILLSTON COBAR BREWARRINA HILLSTON BOURKE IVANHOE WILCANNIA F A J A O D N O S A J J M A M F J CONDOBOLIN WALGETT HAY CONDOBOLIN COBAR BREWARRINA F A J A O D N O S A J J M A M F J 27 PART 3 USING HISTORICAL RAINFAL DATA 1 3 4 8 9 3 3 3 1 5 8 7 6 10 5 52 3 74 1 2 72 9 1 72 0 1 4 3 1 6 33 3 4 58 3 3 3 422 452 559 258 286 487 856 256 103 94 11 2 1 2 Year Year 8 6 2 0 2 6 0 1 5 30 47 63 82 10 17 20 34 41 51 76 10 17 22 3 3 3 33 Dec Dec 206 169 7 9 0 1 6 0 1 5 31 39 52 74 13 21 28 35 47 83 11 16 23 3 33 30 33 229 159 Nov Nov 0 2 6 0 3 5 28 37 50 68 11 17 20 27 34 41 59 10 15 21 30 31 27 26 Oct Oct 220 154 0 1 5 0 1 3 6 0 1 3 19 27 42 60 10 13 18 27 38 51 10 15 2 2 25 30 Sep Sep 169 106 3 0 1 6 9 0 0 2 6 0 1 3 22 26 36 57 97 14 18 25 36 49 99 11 16 2 2 2 22 Aug Aug OFFICE

8 0 2 6 0 1 2 29 35 45 68 13 18 21 28 42 54 11 18 23 30 29 24 26 Jul Jul 153 169 OFFICE POST

4 8 0 0 2 6 9 POST 13 18 23 30 42 52 74 45 58 14 19 27 34 31 33 27 25

Jun Jun 192 128 E ARRINA K W 2 4 2 0 5 0 0 1 3 11 16 24 30 36 51 80 47 62 10 15 21 27 41 3 3 3 30 RE OUR 194 196 May May B B 0 0 2 5 9 0 0 0 1 3 6 13 25 36 56 86 47 77 14 21 35 4 42 31 28 Apr Apr 304 205 5 0 9 1 4 9 0 0 0 3 5 16 23 32 39 63 97 47 10 18 24 37 3 5 3 48 274 101 224 Mar Mar (The State of Queensland, Department of Primary Industries, June 2003) 1 1 6 0 0 1 5 14 21 31 45 60 83 68 94 10 17 28 37 51 4 54 49 49 Feb Feb 123 324 283 3 1 7 4 7 0 0 1 5 7 13 22 30 39 50 80 55 95 13 24 29 40 6 5 3 48 Jan Jan 129 354 206 eviation eviation D D robabilities of monthly rainfall recorded at robabilities of monthly rainfall recorded at tandard tandard tandard tandard P in 100%, 90% … 0% of years or exceeded (mm) received of rain Amounts P of years in 100%, 90% … 0% or exceeded (mm) received of rain Amounts S S Highest on record Mean 20% yrs at least 20% yrs at least 10% yrs at median, 50% yrs at least median, 50% yrs at least 40% yrs at least 30% yrs at 80% yrs at least 80% yrs at least 70% yrs at least 60% yrs at Lowest on record Lowest least 90% yrs at 10% yrs at least 10% yrs at Highest on record Mean 30% yrs at least 30% yrs at least 20% yrs at 60% yrs at least 60% yrs at least median, 50% yrs at least 40% yrs at 90% yrs at least 90% yrs at least 80% yrs at least 70% yrs at Lowest on record Lowest Table 3.1Table Monthly rainfall deciles for 12 locations across the project area. Tables have been reproduced from Rainman Streamflow Version 4.3 USING HISTORICAL RAINFALL DATA PART 3 28 S Mean Highest onrecord 10% yrsat least 20% yrsat least 30% yrsat least 40% yrsat least median, 50%yrsat least 60% yrsat least 70% yrsat least 80% yrsat least 90% yrsat least Lowest onrecord Mean Highest onrecord 10% yrsat least 20% yrsat least 30% yrsat least 40% yrsat least median, 50%yrsat least 60% yrsat least 70% yrsat least 80% yrsat least 90% yrsat least Lowest onrecord S Amounts ofrain (mm)received orexceeded in100%,90%…0%ofyears (1881–1965Cobar PO, 1962–2004Cobar MO) P Amounts ofrain (mm)received orexceeded in100%,90%…0%of years P tandard tandard robabilities of monthly rainfall recorded at at recorded rainfall monthly of robabilities at recorded rainfall monthly of robabilities D D eviation eviation 216 240 Jan Jan 3 47 3 2 97 60 41 31 20 12 72 37 26 16 7 5 3 0 9 5 3 2 0 0 4 7 3 281 116 112 Feb Feb 48 30 24 3 59 46 30 23 14 73 41 25 17 10 8 4 1 0 6 4 1 0 0 9 Mar Mar 238 259 42 3 1 3 74 50 40 28 17 10 57 30 18 13 0 0 7 4 2 1 0 0 5 2 2 9 2 201 219 Apr Apr 27 28 3 1 45 31 20 14 69 44 30 20 16 0 0 7 4 2 1 0 0 8 5 2 5 8 CO B May May 144 RO 27 30 24 2 65 50 38 32 26 15 11 93 59 40 30 21 14 6 3 1 0 8 1 0 9 B 3 AR K EN M 104 149 Jun Jun

24 29 22 2 64 46 36 30 24 17 13 46 37 25 20 15 10 HILL 8 5 2 0 8 4 0 O 1

CO ( M PATTON 102 Jul Jul 22 26 1 1 54 42 31 26 21 17 11 89 39 31 24 20 16 10 POSITE 7 4 2 0 6 2 0 7 9 Aug Aug 114 24 29 1 1 63 49 39 27 23 17 15 10 91 41 29 24 18 17 12

8 6 3 0 4 0 6 9 STREET 105 155 Sep Sep 22 2 2 2 51 35 27 22 16 13 46 33 25 16 13 11 7 4 2 0 9 6 3 0 3 3 0 ) 131 129 Oct Oct 28 24 25 3 10 78 51 43 32 25 20 13 56 44 32 25 16 13 7 2 0 8 4 0 2 Nov Nov 157 122 31 24 3 2 69 54 37 26 22 16 10 48 31 23 16 11 4 2 1 0 5 3 0 8 2 0 158 180 Dec Dec 31 3 3 2 86 58 43 32 26 17 10 61 32 19 15 4 2 0 0 8 3 0 9 6 5 5 1 Year Year 1 1 13 24 1 22 8 8 497 313 272 101 111 254 296 258 242 59 4 3 3 3 31 3 1 57 88 66 8 00 3 3 9 54 9 72 56 5 1 3 1 9 3 4 1 3 9 29 PART 3 USING HISTORICAL RAINFAL DATA 6 0 0 3 1 3 9 1 0 7 9 4 44 1 69 27 6 89 84 57 3 10 0 1 4 45 49 3 1 52 64 3 3 33 3 424 462 579 448 268 333 287 1 22 29 8 2 31 9 Year Year 0 9 0 2 4 7 0 4 8 29 42 67 27 35 55 68 96 10 15 20 15 20 4 3 26 30 Dec Dec 152 197 5 0 3 7 0 5 9 27 38 49 29 35 45 60 79 11 13 18 22 13 20 3 24 25 30 152 127 Nov Nov 6 0 9 0 8 42 63 80 34 44 52 64 88 13 18 23 28 31 16 22 29 3 30 42 33 Oct Oct 150 155 2 1 8 0 9 38 43 62 27 31 35 44 66 11 17 22 29 33 12 14 20 3 31 22 24 Sep Sep 106 121 6 0 0 9 0 8 42 48 61 33 37 46 55 69 15 20 25 29 34 15 21 27 3 2 33 24 108 125 Aug Aug OFFICE

4 1 7 0 6 39 45 59 32 36 42 54 71 13 17 22 28 34 12 19 24 3 31 22 24 Jul Jul 101 111 POST

OFFICE

6 6 3 2 9 2 17 21 26 31 36 47 56 68 10 14 17 24 31 37 43 51 74 3 3 2 26 OLIN Jun Jun 116 125 B POST

Y 6 6 0 7 0 7 13 17 23 31 37 44 53 68 14 18 23 27 36 47 55 76 3 3 29 28 134 138 May May HA CONDO 5 1 6 9 0 2 0 2 6 13 21 27 35 48 72 11 15 23 30 41 54 76 3 4 31 30 Apr Apr 151 256 7 7 2 7 0 0 0 1 4 13 19 24 32 42 71 10 16 22 29 39 58 88 3 3 29 45 200 254 Mar Mar 6 9 2 6 9 0 1 0 2 5 9 17 24 32 43 71 15 24 40 48 68 91 3 3 28 42 Feb Feb 204 255 2 3 7 0 1 0 3 7 11 14 22 33 48 69 14 23 31 38 50 71 96 3 27 45 49 Jan Jan 191 255 eviation eviation D D robabilities of monthly rainfall recorded at robabilities of monthly rainfall recorded at tandard tandard tandard tandard Amounts of rain (mm) received or exceeded in 100%, 90% … 0% of years or exceeded (mm) received of rain Amounts P Amounts of rain (mm) received or exceeded in 100%, 90% … 0% of years in 100%, 90% … 0% or exceeded (mm) received of rain Amounts P Mean 10% yrs at least 10% yrs at Highest on record 40% yrs at least 40% yrs at least 30% yrs at least 20% yrs at 70% yrs at least 70% yrs at least 60% yrs at least median, 50% yrs at 90% yrs at least 90% yrs at least 80% yrs at Lowest on record Lowest Mean 10% yrs at least 10% yrs at Highest on record 40% yrs at least 40% yrs at least 30% yrs at least 20% yrs at 70% yrs at least 70% yrs at least 60% yrs at least median, 50% yrs at 90% yrs at least 90% yrs at least 80% yrs at Lowest on record Lowest S S USING HISTORICAL RAINFALL DATA PART 3 30 S S Mean Highest onrecord 10% yrsat least 20% yrsat least 30% yrsat least 40% yrsat least median, 50%yrsat least 60% yrsat least 70% yrsat least 80% yrsat least 90% yrsat least Lowest onrecord Mean Highest onrecord 10% yrsat least 20% yrsat least 30% yrsat least 40% yrsat least median, 50%yrsat least 60% yrsat least 70% yrsat least 80% yrsat least 90% yrsat least Lowest onrecord Amounts ofrain (mm)received orexceeded in100%,90%…0%ofyears P Amounts ofrain (mm)received orexceeded in100%,90%…0%of years P tandard tandard robabilities of monthly rainfall recorded at at recorded rainfall monthly of robabilities at recorded rainfall monthly of robabilities D D eviation eviation 210 213 Jan Jan 31 29 4 3 83 52 31 24 17 13 80 48 30 18 14 1 0 8 4 1 0 8 5 2 1 9 184 217 Feb Feb 28 27 3 3 73 43 30 21 17 12 76 46 32 19 11 0 0 5 2 0 0 8 5 2 5 5 Mar Mar 218 187 28 3 3 4 88 59 39 24 16 11 73 39 27 21 16 10 6 2 0 0 5 4 0 0 2 7 1 182 179 Apr Apr 28 26 1 3 46 28 22 17 11 65 47 35 26 16 13 8 4 2 0 0 9 4 1 0 9 2 IVANHOE HILLSTON May May 124 106 24 27 26 33 20 16 10 59 43 37 31 22 14 10 74 56 45 35 26 3 0 7 1 0 144 123 Jun Jun 22 26 24 3 25 18 11 57 42 31 27 22 18 13 62 54 46 38 31 6 1 9 4 0

4 POST

POST

108 103 Jul Jul 22 30 1 2 21 16 10 50 35 27 23 17 14 11 61 50 39 32 27 OFFICE

5 0 6 3 0 OFFICE 9 3 Aug Aug 24 31 1 2 25 18 13 71 53 41 30 25 21 17 11 84 60 50 42 36 30 6 0 7 3 0 8 0 151 Sep Sep 22 28 2 2 19 16 11 95 49 31 25 21 18 13 49 37 32 28 24 8 0 9 7 4 0 0 1 132 128 Oct Oct 25 28 28 3 24 16 12 63 45 35 27 20 18 13 10 70 55 46 36 31 6 0 5 0 6 Nov Nov 120 149 25 27 27 2 13 58 41 29 20 15 63 48 34 28 20 8 5 3 0 9 6 3 1 0 3 151 107 Dec Dec 27 27 29 3 15 10 74 47 33 18 13 11 73 52 40 27 19 6 3 0 7 3 1 0 2 Year Year 47 3 3 3 29 30 257 1 2 2 1 88 2 10 7 330 264 475 30 5 42 3 3 1 3 69 78 49 2 29 3 0 6 1 1 33 9 58 26 7 1 3 0 5 2 8 3 0 8 2 2 5 1 3 0 31 PART 3 USING HISTORICAL RAINFAL DATA 9 3 3 1 3 4 8 1 1 1 59 0 01 22 9 29 65 0 72 26 46 65 48 97 1 6 67 2 22 28 3 3 1 33 3 4 49 52 554 476 757 228 267 1 92 1 1 1 Year Year 1 1 4 0 4 0 0 2 4 7 54 67 81 11 18 25 37 58 11 17 25 35 43 2 4 3 26 Dec Dec 201 124 5 9 6 9 8 0 4 0 0 1 3 5 49 59 79 12 16 27 46 87 11 16 23 31 38 1 3 3 1 177 Nov Nov 8 8 5 9 0 4 0 0 2 3 5 42 58 82 14 20 34 52 13 20 25 29 35 1 3 3 22 Oct Oct 231 110 2 9 6 9 0 2 6 0 0 0 1 3 32 42 59 14 21 33 10 14 19 23 1 1 27 28 Sep Sep 191 140 6 8 0 2 5 0 0 1 2 4 32 42 66 14 22 28 59 12 17 23 27 29 28 11 13 137 Aug Aug OFFICE

6 1 9 0 3 8 0 0 0 2 4 OFFICE 44 56 73 11 18 30 44 92 13 18 22 32 1 2 33 31 Jul Jul

178 POST

POST 5 7 1

0 5 0 0 1 3 5 8 10 15 20 28 34 43 59 82 13 21 30 42 3 1 2 29 Jun Jun 125 127 URRA B OO 8 0 8 3 7 0 1 0 0 0 3 5 9 ALGETT B 14 22 31 41 53 64 95 95 15 25 33 45 3 4 1 2 182 May May W TI 0 4 6 4 8 0 1 0 0 0 0 2 5 8 14 23 30 40 53 87 16 31 58 4 3 1 25 Apr Apr 233 125 0 0 0 5 0 1 0 0 0 1 3 6 11 19 28 43 52 65 90 11 18 38 61 5 4 4 24 216 398 Mar Mar 3 0 5 0 0 1 3 7 11 18 26 38 50 63 95 12 21 26 45 79 4 55 56 29 Feb Feb 139 239 178 0 1 0 0 0 1 3 5 9 17 27 34 45 64 76 98 10 15 29 45 78 64 29 6 5 Jan Jan 148 345 385 eviation eviation D D robabilities of monthly rainfall recorded at robabilities of monthly rainfall recorded at tandard tandard tandard tandard Amounts of rain (mm) received or exceeded in 100%, 90% … 0% of years or exceeded (mm) received of rain Amounts P Amounts of rain (mm) received or exceeded in 100%, 90% … 0% of years in 100%, 90% … 0% or exceeded (mm) received of rain Amounts P Mean 10% yrs at least 10% yrs at Highest on record 40% yrs at least 40% yrs at least 30% yrs at least 20% yrs at 70% yrs at least 70% yrs at least 60% yrs at least median, 50% yrs at 90% yrs at least 90% yrs at least 80% yrs at Lowest on record Lowest Mean 10% yrs at least 10% yrs at Highest on record 40% yrs at least 40% yrs at least 30% yrs at least 20% yrs at 70% yrs at least 70% yrs at least 60% yrs at least median, 50% yrs at 90% yrs at least 90% yrs at least 80% yrs at Lowest on record Lowest S S 3 USING HISTORICAL RAINFALL DATA PART 3 2 S S 10% yrsat least 20% yrsat least 30% yrsat least 40% yrsat least median, 50%yrsat least 60% yrsat least 70% yrsat least 80% yrsat least 90% yrsat least Lowest onrecord Mean Highest onrecord 10% yrsat least 20% yrsat least 30% yrsat least 40% yrsat least median, 50%yrsat least 60% yrsat least 70% yrsat least 80% yrsat least 90% yrsat least Lowest onrecord Mean Highest onrecord Amounts ofrain (mm)received orexceeded in100%,90%…0%ofyears P Amounts ofrain (mm)received orexceeded in100%,90%…0%of years P tandard tandard robabilities of monthly rainfall recorded at at recorded rainfall monthly of robabilities at recorded rainfall monthly of robabilities D D eviation eviation 106 236 Jan Jan 25 25 4 2 60 34 25 17 11 68 39 23 15 11 0 0 8 4 2 0 0 6 3 1 0 1 222 185 Feb Feb 22 26 3 3 57 39 23 16 80 43 28 20 14 6 4 1 0 0 9 6 3 1 0 0 2 4 Mar Mar 145 325 26 1 2 4 45 34 21 16 53 30 20 15 10 9 5 3 1 0 0 6 4 1 0 0 9 3 3 181 118 Apr Apr 25 28 2 1 53 27 19 13 39 26 21 16 12 0 0 7 5 2 0 0 7 4 2 1 1 8 W W May May 113 150 22 24 25 28 17 11 53 43 34 25 19 13 69 43 34 29 24 ILCANNIA ENT 7 4 0 9 4 0 0 W ORTH 102 Jun Jun 22 27 2 2 18 13 49 36 27 22 17 11 90 61 38 34 30 25 9 4 0 8 5 2 0 0 1

POST

POST Jul Jul 24 1 1 1 17 13 11 84 44 29 22 18 14 10 85 48 35 31 27 21 6 1 6 3 1 0 8 7 6

OFFICE

OFFICE Aug Aug 26 1 1 1 18 14 84 43 28 22 18 13 10 92 50 42 35 31 23 9 5 0 6 3 1 0 8 8 8 109 Sep Sep 27 1 1 2 19 14 86 38 30 18 14 10 55 38 31 27 22 9 6 0 7 5 3 2 0 6 7 1 126 121 Oct Oct 25 26 25 28 17 12 10 67 42 30 21 18 14 10 62 42 32 27 20 4 0 5 1 0 Nov Nov 107 22 24 2 2 12 95 48 33 26 19 12 56 40 28 20 15 9 5 2 0 8 4 2 0 0 0 3 252 198 Dec Dec 26 27 4 2 71 44 28 15 11 52 33 25 16 12 8 6 3 1 0 7 4 2 0 0 0 1 Year Year 24 2 1 10 2 1 1 68 7 262 298 275 133 100 287 301 278 4 3 25 3 3 3 26 11 67 1 82 13 8 66 0 0 54 95 47 2 0 5 4 3 0 5 7 1 0 1 7 33 PART 3 USING HISTORICAL RAINFAL DATA essential for sustainable business practice. Informed analysis and use of historical data will be an essential component of this management. considered when making long-term plans (ten years or more) as future temperature and rainfall conditions may vary from the ranges experienced in the past. Nevertheless, for the moment, the historical record provides the best means we have of describing the climatic characteristics of a location. Managing the impact of climatic variability will continue to be The result has been rapidly increasing temperatures and, as discussed in Part 1, temperature strongly influences the processes that create rainfall, as well as evaporation and the biological activity of plants and animals.This means that climate change needs to be There is increasing evidence that the global climate is in a state of rapid change, influenced gassesstrongly‘greenhouse’ produced by by human activity.These changes are likely to continue for at least the next 100 years. A comment on climate change 3 USING HISTORICAL RAINFALL DATA PART 3 4 PART 4 – SEASONAL RISK ASSESSMENT

• The SOI Phase system is the best currently available for western NSW but it is more reliable in winter and spring than at other times FOREWORD • The SOI Phase system is more reliable for (simulated) pasture growth – essentially ‘effective rainfall’ – than for actual rainfall • It’s worthwhile checking the SOI phase in June, July, August and September in particular. Enter the ‘LongPaddock’ website http://www.LongPaddock.qld.gov. au/RainfallAndPastureGrowth/NSW as a favourite in your browser to ensure that you know the current SOI phase (click on SOI Phase). • Probabilities of exceeding median pasture growth will tend to be low in phase 1, and high in phase 2; probabilities around 50% don’t mean you should expect average conditions, just that the SOI phase is not telling you much about likely future conditions • Maps showing the historical probability of pasture growth exceeding the median for 3 month periods, with lead times up to two months, based on the SOI Phase observed in June, July, August and September. Note that current probabilities for risk assessments with zero lead time are preferable to the historical probabilities and maybe accessed through the LongPaddock site http://www.LongPaddock.qld.gov.au/RainfallAndPastureGrowth/ (follow the links to NSW/PastureGrowth SeasonalProbability)

Background In western NSW the SOI phase system provides the best seasonal risk assessments currently available. Testing of this system during the project, and the alternative SST (sea surface temperature) phase system, in terms of the association between the indicator and future rainfall or (modelled) pasture growth based on historical records, supports this conclusion. In undertaking this testing, however, rainfall or pasture growth is not expressed in absolute terms but simply as being above or below the median value for the particular time, place and assessment period. Risk assessments provided by the system are therefore expressed in terms of the probability of exceeding the median. 3 SEASONAL RISK ASSESSMENT PART 4 6 example of the results is shown in Figure 4.2, taken from the model output for Hillston. Hillston. for output model the from taken 4.2, Figure in shown is results the of example An 4.1. Figure in shown are model GRASP the in encapsulated are that relationships The available. are records climate as long as for and – characteristics vegetation and type soil by defined – location particular any for out carried be may calculations These radiation. solar and evaporation temperature, rainfall, of inputs daily from growth pasture daily calculates that Production) GRASP, Grass for (called model simulation computer a from derived were association this test to used data The rainfall. for than growth pasture for greater actually is NSW western in system phase SOI the of strength the Importantly, Source: Queensland Department of Natural Resources and Mines and Resources Natural of Department Queensland Source: model pasture GRASP the of representation Schematic 4.1 Figure the the SeasonalClimateOutlook/SouthernOscillationIndex/SOIDataFiles/ at: found be can numbers SOI detailed More LongPaddock.qld.gov.au/RainfallAndPastureGrowth/NSW at found be can phase current The months. two previous the over SOI in trend the on based month, each of day first the on determined is phase current The SOI the of phases recognised five are There phases SOI grasp P P P SOIL EVAPORATION SOIL hase hase 3 hase 1 hase INFILTRATION DRAINAGE RUNOFF grazing s grazing 5 RAIN y ste Consistently near zero zero near Consistently falling Rapidly negative Consistently m inputs HUMIDITY PASTURE GROWTH PASTURE TRANSPIRATION AVAILABLE SOIL NITROGEN & & WATER NITROGEN SOIL AVAILABLE TEMPERATURE RADIATION & RADIATION P P hase hase hase 4 2 http://www.LongPaddock.qld.gov.au/ UTILISATION Rapidly rising Rapidly positive Consistently CO 2 (click on SOI Phase) SOI on (click http://www. TRANSPIRATION 7

3 PART 4 SEASONAL RISK ASSESSMENT DEC NOV . The. relationship OCT SEP outlook period AUG JUL and the JUN MAY SOI phase period APR than to actual rainfall because calculated pasture growth, taking MAR FEB effective rainfall JAN 0 0

30 45 2

155 150 1

a H kg/ TH W GRO mm ATER W SOIL mm rainfall rainfall for outlook periods that immediately follow the indicator period (i.e. zero lead time) although some useful relationships exists for lead times out to two months. in the winter-spring period, particularly from June to September and cannot be calculated for risk assessments with lead times longer than zero months.The inclusion of some risk assessments with longer lead times in this section is based on other measures of the relationship between SOI phase and pasture growth. All risk assessments based on pasture growth are currently regarded as experimental prototypes.) or spatially uniform to be include Thehere. pattern of variation in the strength of the relationship over the year can be seen in Appendix 1. (Note that when pasture growth is the variable ofscores‘skill’ interest similar to those shown in Appendix 1 At other times the relationship is weaker and/or less coherent over extensive areas.While it may still be useful at particular times or places it is not considered sufficiently strong Generally, the association is strongest: • • also varies depending on the length of the outlook period but this is usually standardised to three months. Unfortunately, the strength of this association, and hence the value of the risk assessment that can be provided, is not uniform across western NSW or at all times of the Ityear. also varies depending on the how far in advance the assessment is made i.e. the lead time or the number of months between the The stronger association with pasture growth means that the SOI phase is more strongly related to into account the factors shown in Figure 4.1, is the best measure of effective rainfall. are derived from the daily output of the model over the entire historical record. On average, pasture growth drops in winter allowing soil moisture to rise which in turn allows rapid growth in spring. Soil moisture drops in late spring to early summer despite high rainfall in October due probably to increased evapotranspiration. Figure 4.2 An example of long-term (1889–2004) monthly averages of rainfall (from the historical record) and soil moisture and pasture growth derived from the GRASP model. Soil moisture and pasture growth averages 3 SEASONAL RISK ASSESSMENT PART 4 8 is entered as a favourite in your browser so you can readily determine the current SOI Phase. SOI current the determine readily can you so browser your in favourite a as entered is site the sure Make 26. Figure by given is October-December period 3-month the in median the exceeding growth pasture of probability historical the June-July), in pattern SOI the on (based August in 1 phase in is SOI the If example: For • • • to corresponding – table growth pasture median exceeding the obtain to used be can figures these how shows Table4.1 figures. these in shown is project the by covered area the Only strong. reasonably is phase SOI and growth pasture between relationship the where periods outlook monthly three those for and SOI the of phase each for – growth pasture median exceeding of probability the – assessments risk seasonal the show section this of end the at 45 to 1 numbered figures The 3. Part in given data climate historical the consult to useful be will it conditions these Under other. the or way one indication strong a providing not is Phase SOI conditions it 50% about is median the exceeding of probability the when that Note is. actually figure the what know to useful still is it like would people most as high as not are order this of probabilities While cent. per 40–60 around often are probabilities the phases three other the In positive). (consistently 2 phase in high and negative) (consistently 1 phase in low be to tends probability The years. all in low, or high, E SOI. the in indirectly, albeit reflected, be will turn in which waters, adjacent and Sea Coral the in temperatures surface to related be will so do to capacity NSW. Their western into moisture feeding systems major the are highs the Tasman spring and winter In 1. Part in discussed moisture of sources major the consider we when understandable is spring and winter in association stronger The ven in winter and spring the probability of exceeding median pasture growth will not be be not will growth pasture median exceeding of probability the spring and winter in ven the current SOI phase. SOI current the and interest of period outlook 3-month the month current the . Pasture growth could be well above or well below the median – it’s just that the the that it’sjust – median the below well or above well be could growth Pasture . http://www.LongPaddock.qld.gov.au/RainfallAndPastureGrowth/NSW . Simply look up the figure – given by the number in the the in number the by given – figure the up look Simply . does not mean that we can expect average average expect can we that mean not does historical probability of of probability historical

3 PART 4 SEASONAL RISK ASSESSMENT rapidly rising consistantly near zero consistently negative consistently positive rapidly falling

– – – – –

4 5 SOI PHASE 1 2 3

5 5 20 25 30 35 40 45 10 15

4 9 4 19 24 29 34 39 44 14

3 8 3 18 23 28 33 38 43 13 PHASE

SOI Figure The4.3 project area, showing boundaries of Rural Lands Protection Boards http://www.LongPaddock.qld.gov.

2 7 2 12 17 22 27 32 37 42

COONAMBLE

1 6 1 11 16 21 26 31 36 41 WALGETT ) S NYNGAN CONDOBOLIN BREWARRINA onth

m period k J J A J A S N D J A S O A S O S O N S O N O N D O N D COBAR BOURKE HAY (calendar outloo RIVERINA HILLSTON onth BALRANALD WANAARING m er WILCANNIA mb y septe june jul august current WENTWORTH BROKEN HILL MILPARINKA within the project area, as shown in Figure 4.3. obtain comparable information for rainfall refer to au/SeasonalClimateOutlook/RainfallProbability/ In figures 1 to 45 the boundaries correspond to the Rural Lands Protection Boards included Similar information for rainfall is not presented in this guide because, as explained earlier, the relationship between SOI phase and rainfall is weaker than for pasture growth although the variation in the relationship throughout the year is similar. If you wish to Table 4.1Table Index of figures showing the probability of exceeding median pasture growth for various outlook periods and SOI phases. 4 SEASONAL RISK ASSESSMENT PART 4 0 the historical probabilities shown here are the best available. available. best the are here shown probabilities historical the No NSW/PastureGrowth_SeasonalProbability_SkillScore to link the following by probabilities these with associated skill the see You can to links the Follow monthly. updated are and at found be can They for figures historical the to preferable are probabilities These year. current the as phase SOI same the having record historical the in years those of each from data rainfall actual using date, present the from forward model GRASP the running by conditions), growth current for accounting (i.e. the calculate to possible is it time, lead zero with assessments For probabilities. historical the interpreting when effects these for allowance some make should Landholders present. already is growth of flush good a if figure historical the than higher be may and spell, dry prolonged a following figure historical the than lower be may growth median exceeding of probability the particular, In period. the of start the at growth plant for available moisture soil of amount the and hand on actually pasture of amount the by influenced also is months three next the for growth pasture the indicate figures following the Although note Important current probabilities current , based on long-term growth simulation, the actual pasture growth outlook outlook growth pasture actual the simulation, growth long-term on based , http://www.LongPaddock.qld.gov.au/RainfallAndPastureGrowth/ are available for 1 or 2 month lead times, and for these assessments assessments these for and times, lead month 2 or 1 for available are historical probability of exceeding median median exceeding of probability historical NSW/PastureGrowth SeasonalProbability NSW/PastureGrowth . zero lead time lead zero current probabilities current assessments.

, .

4 PART 4 SEASONAL RISK ASSESSMENT June June, July, August 1 June June, July, August 2 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 1 2 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. SEASONAL RISK ASSESSMENT PART 4 42 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 4 3 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 4 August July, June, June 3 August July, June, June 3

4 PART 4 SEASONAL RISK ASSESSMENT June June, July, August 5 July July, August, September 1 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 5 6 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 44 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 8 7 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 3 September August, July, July 2 September August, July, July 45 PART 4 SEASONAL RISK ASSESSMENT July July, August, September 4 July July, August, September 5 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 9 10 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 46 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 12 11 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 2 October September, August, July 1 October September, August, July 47 PART 4 SEASONAL RISK ASSESSMENT July August, September, October 3 July August, September, October 4 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 13 14 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 48 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 16 15 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 1 October September, August, August 5 October September, August, July 49 PART 4 SEASONAL RISK ASSESSMENT August August, September, October 2 August August, September, October 3 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 17 18 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P 5 SEASONAL RISK ASSESSMENT PART 4 0 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 20 19 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 5 October September, August, August 4 October September, August, August 1

5 PART 4 SEASONAL RISK ASSESSMENT August September, October, November 1 August September, October, November 2 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 21 22 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 52 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 24 23 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 4 November October, September, August 3 November October, September, August 3

5 PART 4 SEASONAL RISK ASSESSMENT August September, October, November 5 August October, November, December 1 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 25 26 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 54 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 28 27 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 3 December November, October, August 2 December November, October, August 55 PART 4 SEASONAL RISK ASSESSMENT August October, November, December 4 August October, November, December 5 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 29 30 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 56 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 32 31 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 2 November October, September, September 1 November October, September, September 57 PART 4 SEASONAL RISK ASSESSMENT September September, October, November 3 September September, October, November 4 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 33 34 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 58 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 36 35 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 1 December November, October, September 5 November October, September, September 59 PART 4 SEASONAL RISK ASSESSMENT September October, November, December 2 September October, November, December 3 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 37 38 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P 6 SEASONAL RISK ASSESSMENT PART 4 0 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 40 39 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 5 December November, October, September 4 December November, October, September 1

6 PART 4 SEASONAL RISK ASSESSMENT September November, December, January 1 September November, December, January 2 period period month month

rowth phase phase utlook utlook urrent urrent C O SOI C O SOI G 41 42 igure igure F F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

30– 0–30 2 0 2 10– robability 0–10 xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) P E produced for the LandWater andWool Managing ClimateVariability Project. rowth G asture P edian M xceeding E of robability P SEASONAL RISK ASSESSMENT PART 4 62 produced for the Land Water and Wool Managing Climate Variability Project. Variability Climate Managing Wool and Water Land the for produced E P xperimental prototypes from AussieGRASS (www.LongPaddock.qld.gov.au) (www.LongPaddock.qld.gov.au) AussieGRASS from prototypes xperimental 0–10 Y T I L I B A B O R 10– 2 0 2 0–30 30–

4 F O 0 4 E 0– 6 0 CEEDING N I D E E XC 6 0– 7 0 7 0– 8 0 8 0– M 9 0 N A I D E 9 0–100 seasonally low low seasonally growth P E R U T S A F F igure igure 44 43 G SOI SOI O C SOI O C urrent urrent utlook utlook phase phase H T W O R

month month period period 4 January December, November, September 3 January December, November, September 3

6 PART 4 SEASONAL RISK ASSESSMENT September November, December, January 5 period month

rowth phase utlook urrent C O SOI G 45 igure F asture P growth seasonally low 0–100 9 edian 0 9 M 0– 8 0 8 0– 7 0 7 0– 6 xceeding 0 6 0– E 4 0 of 4

• Trigger points are calendar dates beyond which decisions to buy or sell livestock should not be delayed FOREWORD • They can be identified by summarising the long-term (simulated) pasture growth record to show when the prospects for pasture growth for the next three months are highest or lowest, and the variability in growth from year to year • Pasture growth profiles and country type descriptions for 27 locations across western NSW based on feedback from graziers • How to use this information to identify trigger points for your own property and to complement or enhance seasonal risk assessments based on SOI Phase

Background Making management decisions that involve taking a chance on future climatic conditions is always difficult – and more so when rainfall is not strongly seasonal, as in western NSW. Understanding the seasonal risk assessments provided by the SOI phase system will be useful during winter and spring but at other times of the year this system is not very helpful. Having a rule of thumb about how long destocking decisions can reasonably be delayed in the hope that the season may improve should assist with these difficult decisions; alternatively, a rule of thumb about when might be the best time to buy if the season already looks promising might assist in getting the best productivity consistent with prudent risk management. In this section we have tried to encapsulate this idea of ‘rules of thumb’ in the form of ‘trigger points’. Trigger points are times of the year – calendar dates – when the prospects of future pasture growth are high or low based on the long-term record. Understanding this long-term growth pattern should be useful, when combined with first hand knowledge of the current season, in deciding whether its time to buy or sell. If in doubt, a decision to sell should not be delayed beyond the autumn trigger point and a decision to buy, while probably less critical, beyond the spring trigger point. TRIGGER POINTS PART 5 66 the 50th percentile (median) values (or the growth potential index) alone. index) potential growth the (or values (median) percentile 50th the from identified be otherwise would that points trigger the of adjustment some to lead may information This values. percentile 80th and 20th the between difference the by indicated – date starting each with associated growth pasture of variability historical the of indication an providing of advantage the has percentiles critical using points trigger Defining another. over date one choose to reason strong no be may there dates, starting of number a for similar are values index the where instances, some in although identified easily are values lowest and highest the as index potential growth the for forward straight is information this from points trigger Defining • • step from derived data raw the summarised have we section this of end the at figures the In points. starting 26 the among identified be to – points trigger i.e. – values minimum and maximum allows that way a in point starting each for data of set the Summarise 2. periods month 3 in produced growth of amount the record, the of year each for Calculate, 1. process: following the by points trigger identify can we information, this From section. previous the in described probabilities historical the produce to used growth pasture daily simulated of record years) 100 (> long-term same the requires points trigger Identifying or the upper limits of the 2nd, 5th and 8th deciles). 8th and 5th 2nd, the of limits upper the or values, historical of respectively 80% and 50% 20%, lowest the define that growth of levels the (i.e. percentiles 80th and (median) 50th 20th, the to correspond that values growth percentiles Critical similar). be would year the over values index of pattern the but point starting each for growth average the using simply over advantages certain has index this (Using growth. low generally of periods commence that those for low and periods, growth high generally with associated dates starting those for high be will index, the of value the thus and area, This growth. pasture of level given any exceeding of probability the gives that curve the i.e. curve probability cumulative the under area the to equal is index the Technically, figure. single a as date starting index potential Growth 1 starting periods month 3 for i.e. year the throughout intervals fortnightly at starting

January, 15 January, 29 January etc. in each year; each in etc. January 29 January, 15 January,

1 in two ways: two in 1 – this approach summarises the historical data by specifying the the specifying by data historical the summarises approach this – – this index summarises the historical growth data for each each for data growth historical the summarises index this – 67 PART 5 TRIGGER POINTS Figure 5.1 Location diagram for trigger point analysis. Numbers in the diagram refer to numbers 5.1Table in and the pasture growth profiles in Figure 5.3 (at the end of this section). 6 WALGETT 7 4 5 CONDOBOLIN 15 3 14 16 8 13 BREWARRINA 24 COBAR HAY 25 BOURKE 9 HILLSTON 26 27 2 21 IVANHOE 23 22 1 WILCANNIA 12 17 19 11 BROKEN HILL 18 20 10 DARETON TIBOOBURRA a modified pattern. profiles.Their choice has been allowed to stand. In a few cases, graziers thought that none of the alternatives was quite right and they adjusted the profiles to better reflect their understanding of the local situation.These profiles have also been included but identified as climatic record that were used to generate these profiles are given in Appendix The2). type of country on each property is described in the table. In some instances, graziers in relatively close proximity selected somewhat different participated in this analysis.They correspond to the numbers 5.1Table in and to the numbers in Figure 5.3, at the end of this section, which show the profiles that graziers felt best described the local pasture growth pattern. (The combinations of model version and advice on which version, if providedany, a good description of the local pasture growth pattern. The numbers shown in the location diagram (Figure 5.1) represent the properties that particular locations based on experimental dataTo or represent‘average calibration’. an identify the best profile for particular locations, we provided graziers across the region with growth potential and critical percentile profiles from likely alternative versions, based on climatic data from the most appropriate long-term weather station. thenWe sought their Identifying trigger points in the way we have described depends on having pasture growth profiles that are acceptably accurate for any particular location. Several different versions of the GRASP model are available for westernThese NSW. have either been calibrated for Identifying trigger points for your property TRIGGER POINTS PART 5 68 local conditions. local of perception co-operator’s the reflect to modified was pattern growth modelled the that indicate entries Shaded 5.1. Figure in identified properties the up make that types country the of Descriptions Table5.1 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1 S ite ite NO . Belah-bluebush grass’. 15% ‘plains belah-bluebush, 25% woodlands, riverine 30% box-pine, bimble 30% Box-pine box-pine Bimble mallee 20% box bimble 80% box-pine bimble 15% belah-bluebush, 25% plains, flood 60% grass Mitchell 10% plains, flood 30% bluebush, 60% plains flood local and grass Mitchell bluebush, belah mulga, of areas smaller copperburr, with ridges 25% and slopes saltbush 50% About mallee and belah-bluebush of areas small with box-pine bimble 40% and mulga 50% About mallee 30% and box-pine bimble 70% About box-pine Bimble box-pine Bimble gidgee/brigalow and box/pine bimble with grass Mitchell and plains flood northern of mixture A plains flood Northern grass Mitchell 30% and floodplains northern 30% brigalow, and gidgee 40% box-pine bimble 5% and mulga 15% woodlands, riverine and plains flood northern 80% Mulga Southern riverine woodlands riverine Southern woodlands riverine 10% plains, saltbush 30% grassland, whitetop 60% grassland cottonbush 50% plains, saltbush 50% gum red river and boree rosewood, box, black scattered with plains grassy Open Mallee belah-bluebush 50% mallee, 50% plains Saltbush plains flood saline and Saltbush pine. 5% woodlands, riverine southern 10% plains, saltbush 15% belah-bluebush, 30% mallee, 40% woodlands riverine southern 20% plains, saltbush 20% belah-bluebush, 60% C ountr y t y pe description pe

69 PART 5 TRIGGER POINTS

INDEX TH W GRO PASTURE

1000 800 600 400 200 0

17/12

03/12

19/11 05/11

risk assessment 22/10

Lower skill for SOI-based 08/10

POINT SPRING TRIGGER 24/09

10/09 27/08 1), hold as growth can be expected to decline

PRIMER 13/08 (phase

the 80th and 20th percentile lines. If SOI assessment is

Expected pasture growth for next 3 months is maximal but can be highly variable as shown by the differences between 30/07 favourable (phase 2) consider stock purchase, if unfavourable

16/07 02/07 risk assessment

2) growth profile will tend towards 80th 18/06 Best skill for SOI-based considering options. If SOI is favourable

will tend towards 20th percentile trend line Not yet in the peak growth period, but start percentile trend line. If unfavourable growth

(phase 04/06

21/05 07/05

POINT

TRIGGER 23/04 AUTUMN not delay destocking 09/04 0 0 decisions beyond here

2 8

Expected pasture growth for next 3 months is minimal. Do 26/03

INDEX

TH 12/03

W EDIAN 26/02

GRO PERCENTILE M PERCENTILE

12/02 PRIMER 29/01 risk assessment

Lower skill for SOI-based 15/01 than the trigger point date

feed supply does not improve prepare for a decision no later

Approaching decision time – if xample of how to use the pasture growth profiles and critical percentile values to determine 01/01 E 0

500

1000 2000 1500 kg/ha) ( TH W GRO described there does not fit your property stop now. Consider the growth potential and critical percentile profiles and determine trigger point dates as shown in the example below (Figure 5.2). Note that no critical percentiles can be provided for those locations where graziers altered the original graphs to reflect their assessment of pasture growth pattern. PASTURE Select the site likely to be most similar to your property from the location diagram and the country type descriptions 5.1;Table in generally the closest will probably be the best choice but there may be exceptions.may feelYou that none of these sites is similar to your property in which case you should stop now. Turn to the corresponding graphs in Figure 5.3; again, if you feel the growth pattern point, some time before the trigger point, when preparation for a decision and consideration of options should start. Figure 5.2 trigger points beyond which decisions that depend on future growth should not be delayed. Note‘primer’ the • • • To identifyTo trigger points for your property: 7 TRIGGER POINTS PART 5 0 high growth potential, or to avoid periods when growth is low or highly variable. variable. highly or low is growth when periods avoid to or potential, growth high of periods with coincide to timed be may example, for lambing, and Calving considerations. other by driven often are these though even profiles growth pasture the of examination an from benefit also may decisions strategic decisions, tactical these to addition In system. phase SOI the by provided assessment risk current the by modified be could decision that but – expectation growth highest of period the into going i.e. – point trigger the of time the around made be sensibly could time this at buy to Decisions high. relatively is assessments risk seasonal SOI-based of value the when time a at spring-summer, in occur often will expectation growth highest of period the defines that point trigger the extreme, other the At the point. as ‘primer’ 5.2 Figure in identified time the around point, trigger the before sometime begin should decision a such for Preparation point. trigger the beyond deferred be not should sell to decision a example, for deteriorating, are conditions seasonal If SOI. the by provided is assistance little when time a at decisions, sale particularly decisions, management support can – minimal is months 3 next the for potential growth pasture which after date calendar the i.e. – time this at point trigger a Identifying low. is phase SOI the on based assessments risk seasonal of value the when autumn, and summer late in occur commonly NSW western in growth pasture minimal of Periods making. decision management in assessments risk seasonal profiles SOI-based of use the enhance and complement growth both can points trigger Understanding pasture and points trigger Using 1

7 PART 5 TRIGGER POINTS

INDEX TH W GRO PASTURE INDEX TH W GRO PASTURE

200 0 1000 800 600 400 200 0 1000 800 600 400

17/12 17/12 17/12

03/12 03/12 03/12

19/11 19/11 19/11

05/11 05/11 05/11 starting date

0 0

8 2 22/10 22/10 22/10

INDEX

08/10 08/10 08/10 TH

W 24/09 24/09 24/09 edian

GRO percentile m PERCENTILE 10/09 10/09 10/09

ox-pine

27/08 27/08 le 27/08

mb 13/08 13/08 i 13/08 b

30/07 30/07 5 30/07

16/07 16/07 16/07

02/07 02/07 02/07 ulga and m

18/06 18/06 18/06 5

04/06 04/06 04/06

21/05 21/05 21/05

oodlands, 1

07/05 07/05 07/05 w

23/04 23/04 23/04

09/04 09/04 09/04

26/03 26/03 26/03

, 30% northern floodplains and 30% Mitchell grass

w 12/03 12/03 12/03

26/02 26/02 26/02

rigalo

12/02 12/02 12/02

3 2 1

er er er 29/01 29/01 29/01

mb mb mb

15/01 15/01 15/01

ulga

0% gidgee and 0% northern flood plains and riverine 01/01 01/01 01/01 site nu site nu site nu m 4 8 0 0 0

200 100 300 500 500

1000 1000

W GRO PASTURE INDEX TH kg/ha) W GRO PASTURE ( TH kg/ha) W GRO PASTURE ( TH Figure 5.3 Pasture growth profiles for the 27 sites identified in Figure 5.1 5.1.Table and TRIGGER POINTS PART 5 72

PASTURE GROWTH INDEX PASTURE GROWTH INDEX PASTURE GROWTH (kg/ha) 1000 1500 300 500 400 100 200 200 0 0 0 A N Bi site nu site nu site 01/01 nu site 01/01 01/01

orthern flood plains flood orthern m mb ixture of northern flood plains and Mitchell grass grass Mitchell and plains flood northern of ixture

15/01 le 15/01 15/01 mb mb mb b

29/01 ox-pine 29/01 29/01 er er er er er 6 5 4

12/02 12/02 12/02 26/02 26/02 26/02 12/03 12/03 12/03 26/03 26/03 26/03 09/04 09/04 09/04 23/04 23/04 23/04 07/05 07/05 07/05 21/05 21/05 21/05 04/06 04/06 04/06 18/06 18/06 18/06 w 02/07 02/07 ith 02/07 b i 16/07 16/07 mb 16/07 le le

30/07 30/07 b 30/07 ox/pine and gidgee/ and ox/pine 13/08 13/08 13/08 27/08 27/08 27/08 10/09 10/09 10/09 PERCENTILE m percentile GRO 24/09 24/09 24/09 edian W b TH 08/10 08/10 rigalo 08/10

INDEX

22/10 22/10 22/10 2 8 w 0 0 starting date starting 05/11 05/11 05/11 19/11 19/11 19/11 03/12 03/12 03/12 17/12 17/12 17/12 200 400 600 800 1000 0

PASTURE GROWTH INDEX 3

7 PART 5 TRIGGER POINTS

INDEX TH W GRO PASTURE INDEX TH W GRO PASTURE INDEX TH W GRO PASTURE

400 200 0 0 1000 800 600 400 200 0 1000 800 600 1000 800 600 400 200

17/12 17/12 17/12

03/12 03/12 03/12

19/11 19/11 19/11

05/11 05/11 05/11 starting date

0 0

8 2 22/10 22/10 22/10

INDEX

08/10 08/10 08/10 TH

W 24/09 24/09 24/09 allee edian

GRO percentile m PERCENTILE 10/09 10/09 10/09

27/08 27/08 27/08

ush and

b 13/08 13/08 13/08

lue

b 30/07 30/07 30/07

elah- 16/07 16/07

16/07 b

02/07 02/07 02/07

18/06 18/06 18/06

04/06 04/06 04/06 all areas of

21/05 21/05 21/05

ith s

07/05 07/05 07/05

allee 23/04 23/04 23/04 m

ox-pine

09/04 09/04 09/04 le

mb 26/03 26/03 26/03 i

12/03 12/03 12/03 0% 4

ox-pine and 30% 26/02 26/02 26/02 b

12/02 12/02 12/02

mb ulga and 9 8 7 i

m b

er er er 29/01 29/01 ox-pine 29/01

0% 0%

mb mb mb

5 7

15/01 15/01 le 15/01

out out

01/01 01/01 01/01 site nu site nu site nu Ab Ab Bi 0 0 0

500 500 500

1000 1500 1500 1000 1000 2500 2000 1500 kg/ha) W GRO PASTURE ( TH kg/ha) W GRO PASTURE ( TH W GRO PASTURE kg/ha) ( TH TRIGGER POINTS PART 5 74

PASTURE GROWTH (kg/ha) PASTURE GROWTH (kg/ha) PASTURE GROWTH (kg/ha) 1000 1000 1000 500 500 500 0 0 0 6 6 Ab site nu site nu site site nu site 01/01 01/01 01/01 of areas 0% flood plains, plains, flood 0% 0% out out

15/01 15/01 b 15/01 lue 5 mb mb mb 0% salt 0% b 29/01 29/01 29/01 m er er er er er ush, 30% flood plains, 10% Mitchell grass grass Mitchell 10% plains, flood 30% ush, ulga, ulga, 11 10 1

12/02 2 12/02 12/02

b ush slopes and and slopes ush b elah elah 26/02 25 26/02 26/02 % % b 12/03 12/03 12/03 b elah- lue

26/03 26/03 26/03 b ush, Mitchell grass and local flood plains plains flood local and grass Mitchell ush, b lue

09/04 09/04 09/04 25 b % ridges ridges % 23/04 1 ush, 23/04 23/04

07/05 5 07/05 07/05 % % w b ith copper ith i

21/05 mb 21/05 21/05

04/06 le 04/06 04/06 b 18/06 ox-pine

18/06 18/06 b urr, s urr, 02/07 02/07 02/07 m

16/07 16/07 16/07 aller 30/07 30/07 30/07 13/08 13/08 13/08 27/08 27/08 27/08 10/09 10/09 10/09 PERCENTILE m percentile GRO 24/09 24/09 24/09 edian W

08/10 08/10 08/10 TH

22/10 INDEX 22/10 22/10 2 8 0 0 starting date starting 05/11 05/11 05/11 19/11 19/11 19/11 03/12 03/12 03/12 17/12 17/12 17/12 0 200 400 600 800 1000 0 200 400 600 800 1000 200 400 600 800 1000 0

PASTURE GROWTH INDEX PASTURE GROWTH INDEX PASTURE GROWTH INDEX

75 PART 5 TRIGGER POINTS

INDEX TH W GRO PASTURE INDEX TH W GRO PASTURE

0 1000 800 600 400 200 0 1000 800 600 400 200

17/12

17/12 17/12

03/12

03/12 03/12

19/11

19/11 19/11

05/11

05/11 05/11 starting date

0 0

8 2 22/10 22/10 22/10

INDEX

08/10 08/10 08/10

W 24/09 24/09 24/09

edian

GRO percentile m PERCENTILE 10/09 10/09 10/09

27/08

27/08 27/08

13/08

13/08 13/08

30/07

30/07 30/07

16/07

16/07 16/07

02/07

02/07 02/07

18/06

18/06 18/06

04/06

04/06 04/06

21/05

21/05 21/05

07/05

07/05 07/05

23/04

23/04 23/04

09/04

09/04 09/04

26/03

26/03 26/03

12/03

12/03 12/03

allee

m 26/02

26/02 26/02

12/02

12/02 12/02 5 4 ox

1 1 13

er 29/01 er er 29/01 le 29/01

ox-pine

b mb mb mb

mb 15/01

i 15/01 15/01 le b

0% 01/01 01/01 01/01 site nu Box-pine site nu site nu Bi 8 0 0 0 400 200 600

500 500

1000 1500 1000 W GRO PASTURE INDEX TH kg/ha) W GRO PASTURE ( TH W GRO PASTURE kg/ha) ( TH TRIGGER POINTS PART 5 76

PASTURE GROWTH (kg/ha) PASTURE GROWTH INDEX PASTURE GROWTH (kg/ha) 1000 1000 1500 500 500 200 600 400 0 0 0 6 30% 1 Belah- site nu site nu site 01/01 nu site 01/01 01/01 0% 0% 5 % ‘plains grass’. % ‘plains

15/01 b 15/01 15/01 b elah- i mb b mb mb mb lue

29/01 29/01 29/01 le er er er er er b b b lue ush 1 1 1 ox-pine, 30% riverine riverine 30% ox-pine, 7 6 12/02 8 12/02 12/02

b 26/02 ush, 26/02 26/02 2

12/03 salt 0% 12/03 12/03 26/03 26/03 26/03 b

09/04 plains, ush 09/04 09/04 w

23/04 23/04 23/04 oodlands, 07/05 07/05 07/05 2 21/05 riverine southern 0% 21/05 21/05 25

04/06 04/06 04/06 % b 18/06 18/06 18/06 elah- b

02/07 02/07 02/07 lue b

16/07 16/07 16/07 ush, w

30/07 oodlands 30/07 30/07 13/08 13/08 13/08 27/08 27/08 27/08 10/09 10/09 10/09 PERCENTILE m percentile GRO 24/09 24/09 24/09 edian W

08/10 08/10 08/10 TH

22/10 22/10 INDEX 22/10 2 8 0 0 starting date starting 05/11 05/11 05/11 19/11 19/11 19/11 03/12 03/12 03/12 17/12 17/12 17/12 200 400 600 800 1000 0 200 400 600 800 1000 0

PASTURE GROWTH INDEX PASTURE GROWTH INDEX

77 PART 5 TRIGGER POINTS

INDEX TH W GRO PASTURE INDEX TH W GRO PASTURE

0 1000 800 600 400 200 0 1000 800 600 400 200

17/12 17/12 17/12

03/12 03/12 03/12

19/11 19/11 19/11

05/11 05/11

05/11 starting date

0 0

8 2 22/10 22/10 22/10

INDEX

08/10 08/10 08/10

W 24/09 24/09 24/09

edian

GRO percentile m PERCENTILE 10/09 10/09 10/09

27/08 27/08 27/08

13/08 13/08 13/08

30/07 30/07 30/07

16/07 16/07 16/07

02/07 02/07 02/07

18/06 18/06 18/06

04/06 04/06 04/06

21/05 21/05 21/05

ush plains, 10% southern

b 07/05 07/05 07/05

23/04 23/04

% salt 23/04

09/04 09/04 09/04

ains ush, 1

L 26/03 26/03 b 26/03

lue

b 12/03 12/03 % pine. 12/03

26/02 26/02 elah- 26/02 b

12/02 12/02 12/02 1 0 9 2 2 1

oodlands,

er 29/01 er er 29/01 w 29/01

mb mb mb

allee, 30%

ush plains ush and saline flood p 15/01 15/01 m 15/01 b b

alt alt 0% 01/01 01/01 riverine 01/01 site nu site nu site nu S S 4 0 0 0

400 200 600

500 500

1500 1000 1500 1000

kg/ha) W GRO PASTURE ( TH W GRO PASTURE kg/ha) ( TH W GRO PASTURE INDEX TH TRIGGER POINTS PART 5 78

PASTURE GROWTH (kg/ha) PASTURE GROWTH (kg/ha) PASTURE GROWTH (kg/ha) 1000 1500 1500 1000 1500 500 500 500 0 0 0 5 O m site nu site nu site 01/01 nu site 01/01 01/01 0% 0% pen grass pen allee

15/01 15/01 15/01 m allee, allee, mb mb 29/01 mb 29/01 29/01 er er er er er y plains plains 5 24 22 2 0% 0% 12/02 12/02 3 12/02

26/02 26/02 26/02 elah- w 12/03 scattered ith 12/03 12/03 b lue 26/03 26/03

26/03 b ush 09/04 09/04 09/04 b

23/04 lac 23/04 23/04 k

07/05 07/05 b 07/05 ox, rose ox, 21/05 21/05 21/05 04/06 04/06

w 04/06 18/06 ood, 18/06 18/06 b 02/07 gu red river and oree 02/07 02/07 16/07 16/07 16/07 30/07 30/07 30/07 13/08 13/08 13/08 27/08 27/08 27/08 10/09 m 10/09 10/09 PERCENTILE m percentile GRO 24/09 24/09 24/09 edian W

08/10 08/10 08/10 TH

22/10 22/10 INDEX 22/10 2 8 0 0 starting date starting 05/11 05/11 05/11 19/11 19/11 19/11 03/12 03/12 03/12 17/12 17/12 17/12 200 400 600 800 1000 0 200 400 600 800 1000 0 0 200 400 600 800 1000

PASTURE GROWTH INDEX PASTURE GROWTH INDEX PASTURE GROWTH INDEX

79 PART 5 TRIGGER POINTS

INDEX TH W GRO PASTURE INDEX TH W GRO PASTURE

400 200 0 1000 800 600 400 200 0 1000 800 600

17/12 17/12 17/12

03/12 03/12 03/12

19/11 19/11 19/11

05/11 05/11

05/11 starting date

0 0

8 2 22/10 22/10 22/10

INDEX

08/10 08/10 08/10

W 24/09 24/09 24/09

edian

GRO percentile m PERCENTILE 10/09 10/09 10/09

27/08 27/08 27/08

13/08 13/08 13/08

30/07 30/07 30/07

16/07 16/07 16/07

oodlands

02/07 w 02/07 02/07

18/06 18/06 18/06

04/06 04/06 04/06

21/05 21/05 21/05

07/05 07/05 07/05

23/04 23/04 23/04 ush plains, 10% riverine

ush grassland b

b 09/04 09/04 09/04

26/03 26/03 26/03

12/03 12/03 0% cotton 12/03

oodlands

26/02 w 26/02 26/02

12/02 12/02 25 12/02 27 26

er ush plains, 29/01 er er 29/01 b 29/01

mb mb

hitetop grassland, 30% salt

15/01 15/01 w 15/01

0% salt

outhern riverine 0% 01/01 01/01 site nu 01/01 5 site nu site nu S 6 0 0 0

200 400

500 500

1000 1500 1500 1000

kg/ha) W GRO PASTURE ( TH kg/ha) W GRO PASTURE ( TH W GRO PASTURE INDEX TH 8 TRIGGER POINTS PART 5 0 PART 6 – SOURCES OF CLIMATE AND WEATHER INFORMATION

Where to go for • short term weather information, on the Internet or by fax • seasonal climate outlooks (next 3 months), on the Internet or by fax • historical climate data

As agriculturalists we have rapidly moved from having too little climate and weather information to a problem of information overload. Links provided here will readily source relevant information. The challenge in using the information is to be clear about the uncertainty in short term forecasts and longer-term seasonal outlooks, and to use the information for risk assessment.

Short term weather forecasts The following are some of the most relevant sites (with brief comment on their use and interpretation provided by NSW Department of Primary Industries, Tamworth).

Bureau of Meteorology 4 day forecast http://www.bom.gov.au/products/IDG00074.shtml Note that the hashed areas only indicate a chance of rain, not an expectation. The Bureau of Meteorology forecast information is available in text form at http://www. bom.gov.au/weather/nsw/forecasts.shtml. The notes on the weather are particularly helpful in interpreting the synoptic events.

Australian Weather News forecasts http://www.australianweathernews.com/forecast_OCF.htm This provides a link to a map of Australia showing all the Bureau of Meteorology weather zones, where you can get a 7-day forecast for specific sites within any zone. They all include maximum/minimum temperature forecasts as well as rainfall and some have potential evaporation and sunshine hours. SOURCES OF CLIMTE AND WEATHER INFORMATION PART 6 82 • • • • • • • • interest: special of be may that following the includes It (free). 100 630 1800 on obtained be can directory Faxnumbers Meteorology of Bureau The polling). on instructions for manual fax your Fax(consult Poll on available is information base and forecasts weather of range A fax on forecasts Weather wide. Australia useful forecasts long-term and short both of number large a to links contains site This http://members.ozemail.com.au/~sjhop/c&stw.html Weather Tablelands Southern & Central http://wxmaps.org/pix/aus.vv.html at individually days 6 next the of each for examined be can model same The updated. is information as developments for monitored be should block 6–10 day the in indicated changes and daily updated are Forecasts blocks. day five two of each in Australia in fall could that rain of location and amount the indicates It USA. the in studies Ocean–Land–Atmosphere for Centre the by produced is forecast This http://wxmaps.org/pix/prec7.html IG fax numbers for tailored forecasts for 30 agricultural regions throughout Australia. Australia. throughout regions agricultural 30 for forecasts tailored for numbers fax poll specific provides directory The region. each within amounts and times rain forecast of estimates probability as well as develop to expected are they how and systems weather the of description written forecast, day 4 picture, cloud latest includes – Farmweather site each from pictures radar rain for numbers specific provides directory the rain; regional falling currently of maps – radar Rain 004 935 1902 – chart forecast day 6/7 003 935 1902 – chart forecast day 4/5 728 935 1902 – chart forecast day 2/3 007 935 1902 – chart forecast hr 48 002 935 1902 – forecast day 4 region Australian 252 935 1902 – picture cloud latest and chart) synoptic (the chart analysis Prognostic Level Sea Mean E S COLA forecast for Australia for forecast COLA S

8 PART 6 SOURCES OF CLIMATE AND WEATHER INFORMATION l Niño can have, and the E While short term weather information is provided as forecasts or Southern Oscillation Index & Sea SurfaceTemperature update – 1902 935 432 Australian Drought Statement (by the Bureau of Meteorology) – 1902 935 259 1 & 3 month Australian rainfall maps – 1902 935 262 3 month climate outlook – 1902 935 251 l Niño/La Niña summary and background • • • • Seasonal outlooks on fax Bureau of Meteorology numbersFax directory (free) – 1800 630 100 The directory includes the following that may be of special interest This site provides a series of maps showing areas where conditions are assessed as Drought, Marginal or Satisfactory; updated monthly; maps are available back to June 1998. Current NSW conditions http://www.dpi.nsw.gov.au/reader/drt-area variability as well as the current risk. E http://www.bom.gov.au/climate/enso/ This Bureau of Meteorology site provides a good summary of conditions and why they are important to Australia – worth a look to see what impact median rainfall in Australia over the next three months, based on the SOI phase over the last two need months.You to scroll down map.and select‘Australia’ the http://www.LongPaddock.qld.gov.au/SeasonalClimateOutlook/RainfallProbability/ This site uses historical records of rainfall and SOI to estimate the probability of exceeding on the BoM site. Queensland Centre for Climate Applications This outlook shows the probability of exceeding median rainfall for the next three months. It is issued about the middle of each month and is derived from the pattern of sea surface temperatures in both the Indian and Pacific Oceans.Temperature forecasts are also available Bureau of Meteorology Seasonalhttp://www.bom.gov.au/climate/ahead/rain_ahead.shtml Outlook Seasonal climate outlooks provide as assessment of likely rainfall conditions for the next three months. predictions, seasonal climate outlooks are always stated as probabilities. Seasonal Climate Outlooks SOURCES OF CLIMTE AND WEATHER INFORMATION PART 6 84 12, 18, 24 or 36 months. months. 36 or 24 18, 12, nine six, three, or month week, day, past the over temperature and rainfall of maps excellent site web Meteorology of Bureau The Averages’. in ‘Climatic libraries public most in form printed in available is information same The spreadsheet. a into data the download can you and map the on click Just Australia. in sites 1000 for humidity and temperature min. and max. days, bom.gov.au/climate/map/climate_avgs/clim_avg1.shtml site web Meteorology of Bureau the data, climate past of analysis basic a For climate. local the describe probabilities and medians the and important are events extreme of magnitude and frequency the where plans long-term preparing in useful be can They long. as almost for records have centres Most 1877. from Bourke and 1878 January from records has Pooncarie example For countries. other most of envy the are that records rainfall historical have we Australia, in agriculture of history short the data Despite climate historical Accessing

http://www.bom.gov.au/climate/austmaps/ http://www.bom.gov.au/climate/austmaps/ has average monthly rainfall, rain rain rainfall, monthly average has http://www. has has PART 7 – SOME PRODUCER’S VIEWS

All of the graziers who worked with us on this project have their own experience of managing climate variability and their own philosophies of management. We asked a few to give us a snapshot of their approach, an opinion on the importance of seasonal risk assessment and a response to the information provided by the project.

Hugh McLean, Booligal In general most people’s expectations of weather or climate forecasting would be, when is it going to rain and how much will we receive? I must admit that before becoming involved in the Land, Water & Wool Seasonal Forecasting project in July of 2003 this was also my expectation. I manage, in conjunction with my parents, a family owned merino grazing operation in the Booligal district of south western NSW. In a normal year we would join 4,500 merino ewes to merino rams, however since the 2002 drought this has been reduced to 3,000. The key principle of our grazing system is set stocking of paddocks at conservative rates. The seasonal forecasting project has improved my understanding of two major principles. First, the effect of the Southern Oscillation Index and Sea Surface Temperature on the Australian climate. Second and more importantly is understanding forecast probabilities of achieving median rainfall. Farming like all business enterprises is about taking a risk today to hopefully receive a reward at some point in the future. As in all dry area farming one of our greatest inputs is rainfall. Forecast probabilities allow a risk level to be put on the potential of a future event, therefore allowing for a more meaningful decision to be made. Into the future seasonal forecasting will be used to make decisions more for the upside. Meaning that when there is a forecast probability of achieving above median rainfall of greater than 70% (particularly in the period May to October) greater business risks will be taken. This could mean trading stock, taking on agistment or holding over sale stock to achieve greater per head returns.

Ed Fessey, Brewarrina Good management in grazing industries usually entails the following features: • Some experience and knowledge of the past • Ability to continually access the future availability of feed • The courage to make decisive decisions and not deliberate once the decision has been made SOME PRODUCER’S VIEWS PART 7 86 When I was young, there was no SOI, we made do with fresh cow’s milk! (Graziers need a need (Graziers milk! cow’s fresh with do made we SOI, no was there young, was I When patterns. weather in trend reliable fairly a gives linkage off. This turn to many, how and stock, what deciding when guide good a as it use people some and SOI in interest an Wetake weeks. for feed quality cattle the give can range the on storm A action. drought take to need we before now, options our consider to longer months have we means It drought. the survive us helped bush The problem. a not are weeds Woody acacia. and saltbush old-man dense very and mounds, bush gutters, deeper creeks, sandier with problem, a is Access etc. acacia, prickly bluebush, saltbush, trees, more lot a with bush, to grass from changed has country our years, 50 last the Over scheme. pilot Plus 2000 West the under land conservation is ha 5,000 some Hill, Broken of north country floodout and footslope, Range, Barrier of ha 75,000 on cattle Grey Murray and sheep merino Wegraze rain. useful actual and indications good between involved, distance large the over slip, a many is There direction. westerly a from event trigger a by upon acted is layer moist the when falls rain and coast, the up north well is high a when coast east the from Hill Broken Bevan, Peter luck! good of dose healthy a and making decision planning, good involves management good all day, the of end the At decisions. making about busy get must you then May, of 1st the by occur to likely not is or Day Anzac by occurred not has event rainfall that if Consequently, period. April to January the in event rainfall major a had usually have seasons successful our of all because year the of period critical most the is Day May to Day Anzac Basically numbers. livestock for points trigger making about and NSW western in variability climate managing better on based is project & Wool Water Land The • • • sense of humour). humour). of sense agree or disagree with the trend. trend. the with disagree or agree to deciding in ruthless be and possible as forecasts range long different many as at Look seasons these in grows feed what of experience to relation in seasons of flow and ebb the for feel a get to try and records rainfall local on time Spend sale for ready mob sale next the have always to Plan I have learnt, at the climate forum, that our moisture tends to come come to tends moisture our that forum, climate the at learnt, have I forecast. term short the in made been have strides Great improving. is it but on, rely to enough good is forecasting long-term before go to way long a have we believe I out. pan will season the how of indication enough not have year,and the of times critical at so more closely, weather the watch Australia of area every nearly in graziers and Farmers 87 ACKNOWLEDGEMENTS Wool innovation Limited. Variability Sub-Program of LandWater andWool for his dedication to the cause of improved seasonal risk management over many years and for his guidance throughout this project. Finally we gratefully acknowledge the funding provided by the NSW Department of Primary Industries and the Land,Water andWool partners, Land Australia andWater, and Australian the imagery provided by the Queensland group was undertaken proficiently by Mr Ian McGowen of NSW DPI as part of the analysis phase of the project. expressWe our appreciation to Dr BarryWhite, Coordinator of the Managing Climate led by Dr Beverley Henry, were always extremely obliging and prepared to expend whatever effort was required to undertake the major computing tasks required of the project, answer our questions and provide whatever data and images were required. Further analysis of particularly the 320 wool growers who participated in the project in various ways, many of whom provided valuable feedback in response to surveys or newsletters, or participated in workshops. Our colleagues in the Queensland Department of Natural Resources and Mines, This guide is the result of input from numerous colleagues and collaborators. thankWe ACKNOWLEDGEMENTS GLOSSARY

This glossary contains terms used in the text as well as others likely to be encountered in the literature of climate science and seasonal risk management. Analogue year A year with a similar pattern of the SOI, or some other indicator, to the current year. Anticyclones Cells of high pressure associated with dry air, resulting in mainly cloud-free skies and little or no rainfall. Anticyclones move from west to east across Australia at 25–40°S. Aspect Direction faced (e.g. east or west). Atmosphere The envelope of gases, bound by gravity, which surrounds the earth. Atmospheric pressure The pressure at a point due to the weight of the column of air above it. 88 AussieGRASS Australian Grassland &. Rangeland Assessment by Spatial Simulation. A system operated by the Queensland Centre for Climate Applications that computes pasture growth in 5X5 km grid cells across Australia and provides monthly updates. GLOSSARY Average See mean. Cell A vertical circulation of the atmosphere in which warm air rises and cools, flows laterally at high levels, then descends. Climate The weather experienced by a site or region over a period of many years. Cold front The boundary between a mass of advancing cold air and a mass of warm air. Cyclones Areas of low pressure (depressions) associated with rising warm air and clockwise air circulation in the southern hemisphere. A tropical cyclone is an intense depression fed by very warm (28°C) waters, and by release of latent heat energy as water vapour condenses to form rain. Deciles These divide a set of recorded rainfalls (monthly, seasonal or annual) into 10 equal groups. The lowest 10 per cent of falls belong in decile range one, the next lowest 10 per cent in decile range two and so on up to the highest 10 per cent of recorded falls, which belong to decile range 10. The value between decile range five and decile range six is the median. Decile ranges give a better indication of how dry or wet a period has been than departure from the mean (or average). Drought Meteorological drought occurs when rainfall (for a period of at least three months) falls within the lowest 10 percent of all totals for that period (decile range one). The Bureau of Meteorology calls this a ‘serious deficiency’. Droughts in New South Wales are often associated with strongly negative Southern Oscillation Index (SOI) values, commonly referred to as El Niño events. El Niño Originally this term referred specifically to a warming of the sea surface off the coast of Peru. It is now more generally used to describe the unusual warming of a large area of the eastern equatorial Pacific Ocean. This warming is strongly linked to negative phases of the Southern Oscillation. ENSO El Niño-Southern Oscillation – a composite term referring to the range of events associated with changing SOI phases.

Evapotranspiration Moisture lost to the atmosphere from plants, soil and bodies of 89 water. GLOSSARY GRASP A computer model that calculated daily grass production (hence GRASP) from daily rainfall and other meteorological data. The model has been extensively validated for numerous vegetation types and locations in inland Australia, including western NSW, and is the basis of the AussieGRASS system. Humidity The water vapour content of the air. IPO Interdecadal Pacific Oscillation. The phenomenon is related to sea surface temperature changes on roughly decadal time scales. Indications are that the effects of El Niño and La Niña on Australian rainfall may be modified depending on whether the IPO is in a ‘warm’ or ‘cool’ phase. Isobar A line on a weather map joining places with the same mean sea level barometric pressure. ITCZ The Intertropical Convergence Zone where the moist south-east trade winds of the southern hemisphere meet the north-east trades of the northern hemisphere. It is a zone of heavy rain and thunderstorms, and a main source of tropical rain. JMA Japanese Meteorological Agency. Satellites operated by this agency provide the cloud images of Australia. La Niña This term is now used to refer to the opposite of an El Niño (the opposite phase of the Southern Oscillation). La Niña refers to cooler than average sea surface temperatures in the eastern Pacific which are accompanied by positive SOI values. Mean (average) The mean is calculated by adding all the numbers in a series (e.g. annual, monthly or daily rainfalls) and dividing the total by the number of observations. Median The median value is determined by ranking a set of numbers from highest to lowest. The value which has 50% of the numbers above it and 50% below is the median. Monthly means may be well above the median, especially in arid regions where the mean is distorted by rare, but torrential rainfall events (see mean). MJO Madden-Julian Oscillation, also known as 30–60 day waves, or intra-seasonal oscillations. These periods are low pressure waves that sweep west-east across the north of the continent irregularly every 30–60 days (average 40 days), triggering rainfall events. NCC National Climate Centre of the Bureau of Meteorology located at 90 the Melbourne Head Office (GPO Box 1289K, Melbourne, Victoria 3001). NOAA National Oceanic and Atmospheric Administration. NSW DPI NSW Department of Primary Industries. GLOSSARY Orographic rain Rain caused by terrain standing in the way of moisture-laden air, forcing it to rise and cool resulting in condensation and rain. Precipitation Deposits of water, (rain, ice, snow) reaching earth from the atmosphere. Probability Probability of an event happening can be expressed as a percentage. A probability of 70 per cent means there are 70 chances in 100 (seven in 10) of the event occurring. QDPI Qld Department of Primary Industries. Rainman A computer package developed by the QDPI that provides historical rainfall information for numerous locations across Australia and allows relationships between rainfall and SOI to be determined. The most recent version of the package Rainman Streamflow also provides correlations between SOI and flow of many Australian rivers Relative humidity The water vapour content of the air as a percentage of the amount needed to saturate it at the same temperature. SCO Seasonal Climate Outlook. Issued by climate meteorologists at the National Climate Centre. Skill The extent to which the use of an indicator (e.g. SOI Phase) improves the prediction of a variable (e.g. rainfall or pasture growth) compared with the use of historical data alone. It is related to the strength of the relationship between the variable and the indicator (see also Appendix 1). SOI Phases The five phase of the SOI – consistently negative (1), consistently positive (2), falling (3), rising (4) and neutral (5) – used for seasonal risk assessment. The phases are determined by the pattern of the SOI over the previous two months. SOI Abbreviation for Southern Oscillation Index, the index that measures the strength and sign of the Southern Oscillation. The index is based on a mathematical formula that expresses the standardised difference in air pressure between Darwin and Tahiti. SST Sea Surface Temperature. SSTs play an important role in Australian 91 rainfall variability, particularly the temperatures across the tropical GLOSSARY Pacific Ocean. Trade winds The south-east winds blowing across the southern Pacific Ocean are referred to as trade winds. They weaken in an El Niño event. Trigger Points As used in this booklet, calendar dates that should prompt tactical management decisions when used in conjunction with other observations of seasonal conditions. Troposphere The lowest layer of the Earth’s atmosphere in which our weather occurs. It extends to a height of about 20 km at the equator, and 10–15 km at mid-latitudes. Trough An extension of a low pressure system. Warm episode This refers to a sea surface temperature anomaly in the eastern Pacific; a warm episode is the same as an El Niño and corresponds with a negative phase of the Southern Oscillation. APPENDIX 1 Variation over the year in the relationship between SOI phase and the probability of exceeding median pasture growth.

The strength of the relationship is expressed as a ‘skill’ score calculated using a statistical procedure (LEPS; linear error in parameter space) which measures the improvement in the probability of exceeding median growth obtained by using the SOI phase compared with long-term history. LEPS scores greater than 10 are considered to indicate ‘significant’ skill. All assessments are based on 3-month pasture growth and zero lead time i.e. growth is related to the SOI phase in the two-month period to the end of the indicated month. Months shown in capitals correspond to the ‘current month’ for making an assessment as used in Table 4.1.

JANUARY FEBRUARY MARCH SOI phase – December SOI phase – January SOI phase – February Growth – January to March Growth – February to April Growth – March to May 92 APPENDIX 1

APRIL MAY JUNE SOI phase – March SOI phase – April SOI phase – May Growth – April to June Growth – May to July Growth – June to August

JULY AUGUST SEPTEMBER SOI phase – June SOI phase – July SOI phase – August Growth – July to September Growth – August to October Growth – September to November

OCTOBER NOVEMBER DECEMBER SOI phase – September SOI phase – October SOI phase – November Growth – October to December Growth – November to January Growth – December to February

seasonally <0 0–10 10–15 15–20 20–25 25–30 >30 negligible LEPS Forecast Skill, www.LongPaddock.qld.gov.au APPENDIX 2 Model versions (either the location for which the GRASP model was calibrated or an ‘average’ calibration) and the climate record used to produce the pasture growth profiles included in Part 5.

Site NO. MODEL VERSION – CLIMATE STATION

1 Lake Mere (Louth) – Tibooburra 2 Lake Mere (Louth) – Bourke 3 NSW average – Bourke with grazier modification (Modified graph with pasture growth index only) 4 Lake Mere (Louth) – Walgett 5 Lake Mere (Louth) – Brewarrina with grazier modification (Modified graph with pasture growth index only) 6 NSW average – Walgett with grazier modification (Modified graph with pasture growth index only) 7 Gilruth Plains (Cunnamulla) – Brewarrina 8 NSW average – Cobar 9 NSW average – Cobar 10 NSW average – Wilcannia 93

11 Kinchega (Menindee) – Wilcannia 2 APPENDIX 12 NSW average – Broken Hill 13 NSW average – Condobolin 14 Kinchega (Menindee) – Ivanhoe 15 Kinchega (Menindee) – Condobolin with grazier modification (Modified graph with pasture growth index only) 16 Kinchega (Menindee) – Condobolin 17 Kinchega (Menindee) – Deniliquin with grazier modification (Modified graph with pasture growth index only) 18 NSW average – Broken Hill 19 NSW average – Hay 20 Kinchega (Menindee) – Deniliquin 21 NSW average – Hay 22 Kinchega (Menindee) – Broken Hill 23 NSW average – Hay 24 Kinchega (Menindee) – Hay 25 NSW average – Hay with grazier modification (Modified graph with pasture growth index only) 26 NSW average – Deniliquin 27 NSW average – Hay

For further information or to organise activities related to the material contained in this booklet contact:

Paul Carberry Climate Advisory Officer Tamworth Agricultural Institute Tel: 02 6763 1132 Mobile: 0411139599 Fax: 02 6763 1222 Email: [email protected]

BETTING ON RAIN

MANAGING SEASONAL RISK IN WESTERN NSW

R B HACKER Y ALEMSEGED P M CARBERRY R H BROWNE W J SMITH