Improved wheat yield and production forecasting with a moisture stress index, AVHRR and MODIS data

Abstract

The objective of this study was to improve the current wheat yield and production forecasting system for Western Australia on a LGA basis. PLS regression models including temporal NDVI data from AVHRR and/or MODIS, CR, and/or SI, calculated with the STIN, were developed. Census and survey wheat yield data from the Australian Bureau of Statistics were combined with questionnaire data to construct a full time-series for the years 1991–2005. The accuracy of fortnightly in-season forecasts was evaluated with a leave-year-out procedure from Week 32 onwards. The best model had a mean relative prediction error per LGA (RE) of 10% for yield and 15% for production, compared with RE of 13% for yield and 18% for production for the model based on SI only. For yield there was a decrease in RMSE from below 0.5 t/ha to below 0.3 t/ha in all years. The best multivariate model also had the added feature of being more robust than the model based on SI only, especially in drought years. In-season forecasts were accurate (RE of 10–12% and 15–18% for yield and production, respectively) from Week 34 onwards. Models including AVHRR and MODIS NDVI had comparable errors, providing means for predictions based on MODIS. It is concluded that the multivariate model is a major improvement over the current DAFWA wheat yield forecasting system, providing for accurate in-season wheat yield and production forecasts from the end of August onwards.

Evaluation of the Integrated Canadian Crop Yield Forecaster (ICCYF) model for in-season prediction of crop yield across the Canadian agricultural landscape

“…However, due to over-fitting or biases in the input data, a skilful model at calibration might not give good forecast results once new input data sets are introduced (Qian et al, 2009b;Schut et al, 2009). An improved evaluation approach is to set aside a sub-set of data for validation and use the rest (training data) to calibrate the model.…”

“…Cross validation, particularly leave-one-out-cross-validation (LOOCV), has been proven to be effective for model evaluations with limited sample size (Efron, 1983;Khan et al, 2010), and has been used in similar studies (e.g. Qian et al, 2009b;Schut et al, 2009;Mkhabela et al, 2011). The LOOCV method was adopted in this study to assess the performance of the ICCYF for the three crops at three spatial scales.…”

“…Bullock, 1992;Boken and Shaykewich, 2002;Doraiswamy et al, 2004;Mkhabela et al, 2011;Johnson, 2014). In recent years, functional models have tended to integrate both climate and remote sensing indices into unified yield prediction models (Prasad et al, 2006;De Wit, 2007;Schut et al, 2009;Cammarano, 2009). Despite the long history of the EO based yield forecasting methodologies, it is only in the last decade or so, when EO datasets became available in near real time (NRT) from open sources (e.g.…”

Yield estimation and forecasting for winter wheat in Hungary using time series of MODIS data

“…Others work with direct mathematical relation between the satellite data and the yield of the given plant, where some of them also include some basic meteorological or agronomical data into the relationship (e.g. Maselli et al 1992;Hamar et al 1996;Schut et al 2009;López-Lozano et al 2015). The procedures are applied for various species, for example, for corn (e.g.…”

Ground truthing protocols for biomass estimation in rangeland environments

“…Recent studies in remote and proximal sensing-based biomass estimation methods and agro-meteorologic models which make use of plant growth and climate variables in combination with remote sensing have shown potential [2], [3]. Remote sensing based approaches offer an advantage in that they are spatially explicit and dynamic but require calibration with ground data [6]. In order for remote sensing based approaches to be effective, ground truthing methods to acquire reference data have to accurately measure the ground conditions the satellite is observing [5].…”