Precipitation forecasting utility for proactive agroecosystem management: A case study from the Texas Gulf LTAR site

Le résumé fourni par la source

Precipitation is a primary driver of plant production across most agroecosystems. As such, enhanced precipitation forecasting will directly benefit agroecosystem management planning and decisions related to stocking rates, crop selection, restoration seedings, and wildfire fuel loads. When precipitation occurs primarily through convective storms, as is the case in the southern Great Plains, precipitation can be extremely dynamic and difficult to forecast. Providing stakeholders accurate estimates of forecast utility would provide them with the information necessary for proactive land management decision-making. In this case study at the Texas Gulf Long-Term Agroecosystem Research site, the objectives were to (1) optimize individual forecast models acquired from National Oceanic and Atmospheric Administration's North American Multi-Model Ensemble and (2) assess the correlation between historical precipitation and retrospectively forecasted (hindcast) precipitation from 1982 to 2024. Optimal forecasts were often obtained from three to five aggregate combinations of individual models that were accurate for most of the year. Seasonal forecasts always produced greater forecast skill than monthly forecasts, and the best forecasts were produced in autumn and spring. Alternatively, when forecasts were acquired in autumn, forecast skill was poorer compared to when forecasts were acquired in early spring, and there were no reliable forecasts produced from 2-month lead times at this site. El Niño Southern Oscillation conditions were, however, accurately accounted for by individual forecast models. Collectively, these findings suggest that there is utility in using precipitation forecasting for proactive agroecosystem management and planning even in dynamic precipitation regions such as the southern Great Plains.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.