Jaclyn Brown1, Chris Sharman2, Dean Holzworth3, Philip J. Smethurst4, Javier Navarro Garcia5, Garry Hopwood6
1 CSIRO Agriculture and Food, 15 College Rd, Sandy Bay TAS 7005, Australia; Jaci.Brown@csiro.au
2 CSIRO, Data61, Sandy Bay TAS 7005, Australia;
3 CSIRO, Agriculture and Food, Toowoomba , QLD
4 CSIRO, Land and Water, Sandy Bay TAS, Australia
5 CSIRO, Agriculture and Food, St Lucia, QLD 6 CSIRO, Agriculture and Food, Black Mountain ACT.
Crop yield forecasts are commonly based on models using historical climate, i.e. climatology. With increasing skill of seasonal climate models, these models are now becoming a feasible alternative to improve predictability. To facilitate this change of practice we have developed a web based tool to test yield prediction skill across climate models and identify agriculturally important problems in the forecasts. Our solution is a cloud based tool AgScore that is called from an R or Python session from which an ensemble of forecasts is uploaded for a location and crop (chosen from a broad, international predefined set). APSIM is executed with the uploaded climate model data, and AgScore analyses the results against identical APSIM simulations using baseline climatology for the same period. A standard suite of metrics are then sent back to the user giving them an indication of climate model performance. We are keen to hear feedback on how to develop this metric to best meet the needs of the agricultural and climate science communities. More details on the AgScore tool and how to use it can be found at https://research.csiro.au/agscore/.