guest ::
| Home > Publications database > Using SEAS5 Seasonal Weather Forecasts for Regional Crop Yield Prediction in a Land Surface Modelling Approach |
| Home > Publications database > Using SEAS5 Seasonal Weather Forecasts for Regional Crop Yield Prediction in a Land Surface Modelling Approach |
| Conference Presentation (Other) | FZJ-2023-00577 |
; ; ; ; ;
2022
Abstract: Seasonal weather forecasts can provide important information for water resources and agricultural planning. However, their coarse spatial and temporal resolution limit the usage for modelling applications such as crop and land surface models and have hindered their widespread use in such models. In this study, we applied sub-seasonal and seasonal weather forecasts from the latest ECMWF SEAS5 forecasting system in a land surface modelling approach using the Community Land Model version 5.0 (CLM5). Simulations were conducted for multiple years forced with sub-seasonal and seasonal weather forecasts over two different domains, one over the German state of North Rhine-Westphalia characterized by heterogeneous land cover and diverse agricultural land use, the other over the Australian state of Victoria that is dominated by large agricultural fields of mostly rainfed winter grain crops. Our results show that the simulations forced with seasonal and sub-seasonal forecasts were able to reproduce recorded inter-annual trends of crop yield, but the inter-annual variability of crop yields was significantly lower compared to the records. The forecast-forced simulations were able to reproduce the generally higher inter-annual variability in crop yield throughout the Australian domain (approx. 50 % inter-annual variability in recorded yields and 20 % in simulated yields) compared to the German domain (approx. 15 % inter-annual variability in recorded yields and 5 % in simulated yields). Also, sub-seasonal and seasonal simulations reflected the early harvest in the drought year of 2018 throughout the German domain, thus capturing one of the main contributing factors to the low annual crop yield. While general soil moisture trends, such as the European drought in 2018, were reproduced in the results from the sub-seasonal and seasonal experiments, we found systematic biases compared to satellite products that could also be observed in the reference simulations forced with reanalysis weather data. The observed biases in the representation of soil moisture, as well as the relatively low inter-annual variability of simulated crop yield, indicate that the representation of these variables in CLM5 still needs to be improved to increase the model sensitivity to drought stress and other crop stressors (e.g., pests, hail, wind).
|
The record appears in these collections: |
