Global within-season yield anomaly prediction for major crops derived using seasonal forecasts of large-scale climate indices and regional temperature and precipitation

View More View Less
  • 1 Institute for Agro-Environmental Sciences, National Agriculture and Food Research Organization, 3-1-3 Kannondai, Tsukuba 305-8604, Japan
  • 2 Meteorological Research Institute, Japan Meteorological Agency, 1-1 Nagamine, Tsukuba, Ibaraki 305-0052, Japan
  • 3 Global Environment and Marine Department, Japan Meteorological Agency, 1-3-4 Otemachi, Chiyoda-ku, Tokyo 100-8122, Japan
  • 4 Aerological Observatory, Japan Meteorological Agency, 1-2 Nagamine, Tsukuba, Ibaraki 305-0052, Japan
© Get Permissions
Restricted access

Abstract

Weather and climate variability associated with major climate modes is a main driver of interannual yield variability of commodity crops in global cropland areas. A global crop forecasting service that is currently in the test operation phase is based on temperature and precipitation forecasts, while recent literature suggests that crop forecasting services may benefit from the use of climate index forecasts. However, no consistent comparison is available on prediction skill between yield models relying on forecasts from temperature and precipitation and from climate indices. Here, we present a global assessment of 26-year (1983–2008) within-season yield anomaly hindcasts for maize, rice, wheat and soybean derived using different types of statistical yield models. One type of model utilizes temperature and precipitation for individual cropping areas (the TP model type) to represent the current service, whereas the other type relies on large-scale climate indices (the CI model). For the TP models, three specifications with different model complexities are compared. The results show that the CI model is characterized by a small reduction in the skillful area from the reanalysis model to the hindcast model and shows the largest skillful areas for rice and soybean. In the TP models, the skill of the simple model is comparable to that of the more complex models. Our findings suggest that the use of climate index forecasts for global crop forecasting services in addition to temperature and precipitation forecasts likely increases the total number of crops and countries where skillful yield anomaly prediction is feasible.

Corresponding author: TI (iizumit@affrc.go.jp)

Abstract

Weather and climate variability associated with major climate modes is a main driver of interannual yield variability of commodity crops in global cropland areas. A global crop forecasting service that is currently in the test operation phase is based on temperature and precipitation forecasts, while recent literature suggests that crop forecasting services may benefit from the use of climate index forecasts. However, no consistent comparison is available on prediction skill between yield models relying on forecasts from temperature and precipitation and from climate indices. Here, we present a global assessment of 26-year (1983–2008) within-season yield anomaly hindcasts for maize, rice, wheat and soybean derived using different types of statistical yield models. One type of model utilizes temperature and precipitation for individual cropping areas (the TP model type) to represent the current service, whereas the other type relies on large-scale climate indices (the CI model). For the TP models, three specifications with different model complexities are compared. The results show that the CI model is characterized by a small reduction in the skillful area from the reanalysis model to the hindcast model and shows the largest skillful areas for rice and soybean. In the TP models, the skill of the simple model is comparable to that of the more complex models. Our findings suggest that the use of climate index forecasts for global crop forecasting services in addition to temperature and precipitation forecasts likely increases the total number of crops and countries where skillful yield anomaly prediction is feasible.

Corresponding author: TI (iizumit@affrc.go.jp)
Save