Seasonal forecasting of pest population dynamics based on downscaled SEAS5 forecasts

Abstract

Among the varied environmental factors that influence insect life-history, temperature has a relatively profound effect that can be mathematically estimated with non-linear equations. Thus, many models that aim to predict insect-pest population dynamics use meteorological data as input to descriptive functions that predict the development rate, survival and reproduction of pest populations. In a previous study, we developed a temperature-dependent population dynamics model for the global insect-pest Bemisia tabaci, and verified its accuracy under field conditions. In the current work, which focused on Northern Israel, seasonal meteorological forecasts from the ECMWF SEAS5 coupled model were spatially and temporally stochastically downscaled by a weather generator tool using records from ERA5-Land reanalysis and meteorological stations. The local, hourly temperature time series served as input data to a population dynamics model, creating an ensemble of seasonal population forecasts from which probabilistic predictions could be made already at the beginning of the season (which lasts from March to November). Post-hoc evaluation of the seasonal forecast was done using the observed station temperatures as model input. Comparisons to predictions made using climatologic temperatures found the weather generator-based ones much more accurate in predicting the timing of each insect generation, although there was no difference between the two approaches in predicting the population size. Moreover, the weather generator-based predictions highly matched field observations made by pest inspectors during the growing season of 2021. Taken together, our findings indicate that the developed forecasting tool is capable of providing decision makers with the supporting data required for smart seasonal planning and economical- and environmental-driven optimal management of agricultural systems.