Extreme weather events have caused considerable damage to the livelihoods of small farmers around the world. The development of strategies to protect farmers from crop losses caused by adverse weather conditions such as drought and floods has become an important priority for governments and donors, especially given the expected increase in the frequency and intensity of extreme weather events due to climate change in the coming decades.
The integration of crop models and phenological monitoring can help in designing such strategies. Biophysical crop simulation models as well as statistical crop yield prediction models can play a key role in monitoring, and forecasting crop yield conditions depending on the climate and phenological information.
The published work by Afshar et al., (2021) shows that biophysical crop simulation models can be leveraged to generate larger synthetic yield datasets, which can then be used to train weather- or satellite-based crop yield estimation models (Figure 1).
Figure 1: Coupling of biophysical APSIM crop model with a statistical model (random forest) for better prediction of crop yield information
The validation results of using of above-mentioned technique have shown that such a framework can increase the accuracy of crop yield prediction models (at field scale) in the estimation of crop yield information.
Such efforts can guide the attempts to design smart phenology-based index insurance schemes and target yield monitoring resources in smallholder farming environments.
Currently, One direction could be to expand the study to include other crop types and locations to see if the methods and findings can be generalized to other smallholder farming environments. Another direction could be to include more detailed monitoring of crop phenology and weather conditions, to further improve the performance of the statistical crop yield models. Another interesting avenue for the next study would be to improve the performance of the crop model, more detailed data on soil, crop management, and other factors affecting crop growth could be collected and incorporated into the model.
For those who are seeking some kind of collaboration in this field, the below paper would provide beneficial information. Please contact me personally for further discussions.
Afshar, M. H., Foster, T., Higginbottom, T. P., Parkes, B., Hufkens, K., Mansabdar, S., … & Kramer, B. (2021). Improving the performance of index insurance using crop models and phenological monitoring. Remote Sensing, 13(5), 924.