Ground Cover North : Ground Cover 063 August 2006 - North
Pseudo yield maps a cost-effective option A new tool allowing a low-cost rapid assessment of whole-farm yield variability -- pseudo yield mapping -- may help farmers justify investment in precision agriculture technology BY EMMA LEONARD n The cost of precision agriculture (PA) equipment and concern over the cost-benefit of investing in it are the two major reasons why growers are cautious about adopting PA, according to a 2004 GRDC survey. So researchers at the Victorian Department of Primary Industries (DPI) have developed a tool for the low- cost, rapid assessment of whole-farm yield variability, a process referred to as pseudo yield mapping (PYM). Crop yields can vary within a paddock (spatially) and between seasons (temporally). This yield variation can be monitored and mapped annually by a harvester-mounted yield monitor and GPS unit. Yield variation can be due to spatial factors such as soil type, depth and nutrient status, as well as seasonal influences such as rainfall, hard finishes or disease. Building a picture of variation across a single paddock can take many years, especially if a rotation includes a mix of cereals, legumes, oilseeds and pasture or hay crops or fallow. However, many growers are uncertain whether the degree of variability on their property justifies investment in PA technology. Researchers from the Victorian DPI, Dr Peter Fisher and Dr Mohammad Abuzar, are using historical satellite images as a method of estimating the level of variability across paddocks or properties before growers need to invest in capital equipment. The output from the PYM tool is a yield map similar to that obtained from harvester-mounted yield monitors, but the yield variability is modelled from relationships rather than measured. PYM uses historical satellite images available from Landsat satellites. For Australasia, these images go back to 1978. The process relies on growers having recorded a detailed rotation history for each paddock and average paddock yields for as many seasons as possible. Using this information, Landsat images are selected for seasons that have similar crop types. This is important because the images for a pulse or cereal crop can be very different and should not be analysed together. Images are usually selected for a period close to flowering, as this is considered the most accurate time for yield estimation from satellite images. However, Dr Fisher emphasises that work is still needed to see how the crop growth stage at the time of image capture influences the accuracy of analysis. The satellite image data is used to create an index that represents the crop biomass. A global standardisation is applied to ensure that major seasonal and spatial differences are retained and depicted on the final map. The final stage of the analysis is to develop the relationship between the remotely gathered biomass data and the actual average yield recorded by the grower for each crop. The result is an average pseudo yield map for each paddock on a property. The final whole-farm map consists of pixels of data averaged over all the years of available data for each major crop type. The PYM approach has been tested on 12 paddocks in the southern Mallee. Results show the average yield at different locations across these paddocks varies from one tonne per hectare to 2.75t/ha. In the trial, the map produced by the PYM approach agreed closely with the comments of the farmer and showed that some paddocks were generally better yielding than others (Figure 1). It also showed that variability within each paddock fluctuates considerably. Some paddocks had very uniform production, while for others the average yield over time varied from one part of the paddock to another. PA tools such as variable rate technology would provide the greatest cost benefit in such paddocks with greater variability. The map of estimated yield variability can then be used by the grower and agronomy consultants to investigate the causes of yield differences and consider the most appropriate economic remedies. This could include further soil testing or investment in PA technology. Tools to help estimate the returns of investment in PA will be part of the GRDC PA manual to be released this year. PYM provides a further level of information to systems that zone a paddock into high, medium or low. This means that growers can identify paddock variability and prioritise the areas that will provide the best return from more precise agronomic management. This research already demonstrates that the technique has great potential and – because it incorporates grower and satellite information – is far more reliable than analysis of satellite data alone. It is hoped that the tool will now be tested in collaboration with growers, farmer groups and consultants across a wider range of conditions. GRDC Research Code DAV00030 More information: Dr Peter Fisher, 03 5833 5222, Peter.Fisher@dpi.vic.gov.au Precision agriculture GROUND COVER AUGUST 2006 22 ENGINEERED FOR THE BIG COUNTRY For full product range go to www.gason.com.au Phone 03 5352 2151 for your nearest dealer, or email email@example.com RGM/GAS31297GC REAP BEFORE YOU SOW. REAP BEFORE YOU SOW. Order before August 31 and save 8%* *8% Factory Bonus on off-season purchases in July and August, for payment in January 2007. 10% deposit required with order. Only while production slots remain available. Air Seeder HydraTILL® Finance plan available. ScariTILLTM Order a Gason air seeder, ScariTILL or HydraTILL® now and reap an 8%* reward before you even start sowing. We've reduced the price but only while production slots are available - once they're gone, they're gone! Order now and pay on delivery in January 2007 (earlier delivery with payment available if required). Air Seeders • 2200 litres to 15,400 litres • Two or three bins • Ground drive or VRT • Front or rear tow to suit most applications • Frame mounted two and three bins (VRT only) ScariTILLsTM & HydraTILLs® • Rigid pull or floating hitch • Five row conventional till, minimum till or row crop • 178, 200, 229, 250 and 333mm tine spacings (depending on bar selection) • Cutting width from 6m to 18.5m • Tine breakout from 400lbs -- 540lbs (spring release ScariTILL) or 0 -- 300lbs and 300 -- 700lbs (hydraulically adjustable, hydraTILL®) PROUDLY TRUE BLUE Figure 1 MAP OF ESTIMATED AVERAGE YIELD DEVELOPED FROM THE PYM APPROACH (grower comments in red).
Ground Cover 062 June-July 2006 - North
Ground Cover 064 September-October 2006 - North