Mr David Gobbett1, Ms Jackie Ouzman1, Ms Christina Ratcliff1, Mr Rob Bramley1
Introduction: Yield gap analysis assesses the difference between actual crop yield and simulated potential yield to derive insights into factors limiting crop production. Precision Agriculture (PA) protocols were used to determine actual yields for 250 paddocks across Australia, monitored over four years. Actual yield estimates were derived from yield maps – generated from GPS-linked yield data logged by harvester yield monitors.
Methods: The Python workflow implemented a well-established PA protocol for converting raw yield monitor point data into interpolated maps. Key features of the workflow included a module to identify and process different yield monitor data file types; use of open source Python geospatial libraries (e.g. ogr, fiona, shapely) for spatial data operations; a status tracker to monitor the processing of each file.
Results: Over 600 maps were generated with input files averaging 75,000 georeferenced data points (max > 1,000,000 points). The error tracking capability supported timely and focussed interaction with the project agronomic consultants. The highly automated process processed yield maps quickly, but we found manually generating the final GIS map layout was an essential step in ensuring map quality. Agronomic consultants rated 89% of the final maps as ‘OK’ or ‘Excellent’.
Conclusion: The Python workflow made an essential contribution to generating valuable information about crop productivity, with implications for broad acre crop production nationally. Diverse data issues meant that an informed, manual quality control check of each map was essential. The code developed has an ongoing legacy through the open source pyprecag library for PA data analysis.
David Gobbett: As a Senior Experimental Scientist in CSIRO’s Integrated Agricultural Systems program David has led components of several high impact national projects. He contributes specialist expertise in spatial analysis and statistical modelling, development of analytical software tools and technical capability in sensors and instrumentation. Recent involvement in GRDC projects includes Spatial temperature measurement and mapping tools to assist growers, advisors and extension specialists manage frost risk at farm scale; Data fusion, spatial analysis and technical lead in developing the Yield Gap Australia website under GRDC Yield Gap projects.
Jackie Ouzman: Jackie has over 15 years’ experience in integrated agricultural systems, working across many farming systems including; broadacre, sugar, horticulture, viticulture, and livestock. Jackie’s strengths include; modelling of biophysical and economic processes, the management and analysis of complex spatial, biophysical, economic and socio-economic data.