COMMENTS and OUTCOMES

from the

Short Term Climate Forecasting Workshop

BETH EBERT (Bureau of Meteorology) - Climate Forecasting Does it Work?

BOM provides a range of products, at a range of scales and prices.  For example four day forecasts are free to air whilst the Farm Weather is a $3.00 fax service providing prediction of two day rainfall totals and their probabilities.  The DNR SILO facility has a range of products which include Meteorograms for 165 stations with seven day forecasts for rain, temperature and wind.

More sophisticated weather prediction products include GASP (Global Assimilation and Prediction System) which predicts weather up to seven days at a 0.75^o (75 km) resolution and LAPS (Local Area Prediction Model) which is a subset of GASP and has a two day prediction window at 0.375^o (37.5 km) resolution.  Special LAPS runs (36 hour prediction and 5km grid) are available for Sydney and Melbourne.

These specialist models have been assessed using a range of prediction tests including bias score and False Alarm ratio.  Basically error in rainfall prediction increase as time increases to 5-7 days.

Also much easier to predict rain depth in frontal storms rather than convective storms.  Hence Sydney rain is forecast more accurately than Darwin’s rainfall.

The BOM are very interested in providing products that can be used by the farming industry or their specialist advisor? – the Ensemble Prediction System is a variation of the GASP system and there is optimism this can predict 5-7 day rainfall at 75 km grid especially rain total rather than rain for each day is used.  There is sufficient historic data to run model for the last 5 years and check predictions using broadcasting.

BOM would be very interested in a joint project with DNR or DPI to deliver these more sophisticated forecasts to the rural communities.

COMMENTS

• Need to highlight both the economic and environmental benefits from the RWUE initiative.
• Accepts that a 5-10% gain in WUE from STCF makes it well worthwhile.

BRETT TUCKER (NPIRD) - Irrigation Efficiency and Benchmarking A National Perspective

Pointed out that there are trends to reducing the gap between the largest and smallest water users in the dairy industry and that short term climate forecasting (STCF) targets the top 10% of growers.   He indicated that weather research on a farm scale is worthwhile but probably not cost effective for NPIRD and believes that STCF will be complex for farmers to adopt.  Argues that scheduling and technology change are intimately linked.   Believes that STCF works best on water on demand farms, not 4 day prior ordering farms.

COMMENTS

• Can see major potential non-irrigation scheduling benefits from STCF e.g. water-logging, pesticide spraying – but not confirmatory data seems available.
• Many farmers make major farm decisions using non quantitative climate forecasts – such decisions include deferring tractor purchase.
• In southern Australia WUE benefit accrues to the whole community through reduced deep drainage but lower yield risk is carried exclusively by individual farmers.
• Sum of total 5 day rainfall is probably better prediction than individual daily value.
• Are probabilistic predictions useful to the farmer? i.e. X% chance of Ymm over next 5 days.
• There was concern about lost production if irrigation delayed, but predicted rain does not occur.
• Low probabilities of rain and high probabilities of rain most useful to farmers.
• Water logging and crop quality are major issues in Southern NSW.
• Water savings are not high on farmer list of priorities.
• Benchmarking at farm level is important to encourage farmers to lift their game.
• Farmers are using existing forecasting tools, however each user self calibrates to their own acceptable level of risk.

RICHARD WELLS (Grape Grower)

Will STCF help farmers? YES.  Will STCF improve irrigation efficiency?   Wanted to know the likely weather conditions after a rain event.  For example, extended overcast weather rots the grapes.   In below 100% allocation years, the forecasting will be in high demand.  Wanted potential evapotranspiration (ET) prediction for the next 3-5 days - does not have to be highly accurate.

IAN GORDON (DNR)

The environmental benefits may equal or exceed the irrigation water savings.

PAUL DALTON (Consultant)

Evaporation from storage and deep drainage are the major losses on a farm.

IAN HAYLLOR (Cotton Farmer)

Had a strong systems approach, looking for dollars per ML not just yield per ML.  Strongly supports benchmarking especially local benchmarking.  Farm dam storage efficiency is very important to him.  Thought that the driver for WUE adoption is the dollars return.  The industry needs to have confidence in science and its local relevance.

Uses Enviroscan moisture sensors and a consultant who tells when to irrigate.  Can identify most water stressed paddocks as well as the freely draining paddocks.   Irrigates on a 100mm soil water deficit and is trying 85mm next season. 

Farmers like on-farm demonstrations.  Some suspicion of models.  The equipment must be reliable.

Will 7 day STCF help?  It will reduce the chance of water logging and water stress.   The precision of predictions should be high because the cost of delay in watering the crop can be large.   For example an eight day delay in watering costs $1000 per acre in lower yield.  The practical limitations of STCF include the 10-14 days it takes to irrigate the entire farm. With new farms, it can be reduced to 3 days to irrigate the whole farm. 

Strongly supports deeper farm storages (8m+) to reduce the surface area/volume ratio and hence reduce evaporation. This will pay back the cost in 2-3 crops.

COMMENTS

• Got to focus on bottom 40% of growers as this has the potential for maximum gains.
• Use farm field days.
• Wanting to quantify the yield losses of getting behind in watering.
• Irrigation is only part of a whole farm system.
• Economics will drive change.
• Enviroscan allows easy access to soil water information.
• Not sure whether STCF can increase water efficiency, but if you look at all the components of STCF, gains are probably positive.
• Can farm management capture the variability seen by Paul Dalton in his farm irrigation efficiency studies?

QUESTION 1  - Free for all to identify R&D issues considered important by participants

• Better modelling of water logging in crops.
• BOM 5 day forecasts of the probability of total rainfall and total evaporation.
• STCF incorporated into a whole farm management system, not just irrigation scheduling.
• 2D models of furrows and hills.
• Better communication between BOM and farmers.
• The trade off between water stress and water logging needs to be quantified.
• The environmental implications of STCF are important to quantify - the big payoff?
• Ways to efficiently access data sets for modellers.
• Work with farmers to understand and document their decision making process, rather than just focus on a STCF tool.
• What Farmers Want:
• Case study of what a farmer might find more useful in the use of STCF.
• Cotton growers more concerned about water logging rather than water stress.
• Big variation in rainfall across farm during convective storms.
• Want the probability of a rainfall event|amount on a particular day.
• Want to know future evapotranspiration - 3-4 days into the future.
• Only >30mm events in crop may cause water logging - as these past irrigation events are few and far between, water logging not a big problem.  Contrasting view from another farmer.
• Want to know when it won’t rain.
• In horticulture if there is no rain for two weeks - big advantage for planting decision.
• Sugar farmer - too much to lose in yield reduction if get STCF wrong in summer. But probably OK for autumn and winter irrigation as ET much less than 8mm/day.
• Dairy want >4mm rain within 24 hours of spinning nitrogen fertilizer to reduce volatilisation losses.  Up to 8 fertilizer applications a year.
• Horticulture - soil temperatures are important. This is affected by cold fronts. The same issue exists in cotton and rice.  In Cotton cool wet conditions give high seedling mortality.

QUESTION 2 - “What potential advantages do you see if the Short Term Climate Forecasting Irrigation Scheduling system were to be implemented?   Consider also non irrigation payoffs of short term climate forecasting”

• Water logging.
• Pesticide spray schedule.
• Water savings - but magnitude will vary between locations, crop and irrigation set up.
• Response for STCF has been analysed as a mean response. What about the big positive response years.
• Use weather < 7 days period.
• Add stochastic error to hindcasting analysis to gauge positive response.
• Use precious ring tank water to plant - but then rain follows - better climate data could allow better decision making. This outcome would be a very positive response later in the crop cycle.
• Will STCF be an orphan product?  Advances will come only if it gets a big customer base who apply the technique.

Environmental benefit - reduced runoff, salinity implication, deep drainage reduction.

QUESTION 3 - What further research and development needs to be carried out?   “What is the best way to progress this idea?   Where should this happen?”

• Horticulture - wants better service from BOM and their intermediates.
• What are the delivery mechanisms to farmers - electronic, etc.
• Water logging - important but not quantified - implication for sugar and cotton industry.
• Crop water use model for non specialist users.

QUESTION 4 - “Is it possible to confirm the climate-scheduling concept with experimentation?   Where?   By whom?   How should farmers be compensated for allowing the system to be tested on their property?   How should the system be implemented?”

• Relaxed about compensation if treatments  - don’t give a positive response - similar to other on farm experimental studies.

QUESTION 5 - “Can models help devise new management systems?   How do you extend limited experimental evidence to cover wider areas, and other periods of time?” 

• Yes but with difficulty!
• Farm demo sites could be used to run model in real time and use models to work out pay back period for differing irrigation systems (incl. fertilizer response).
• Emphasise verification of models on farm e.g. orchards.