The Answer for an Improved NYS Drought Forecast Is Blowing in the Winds

Drought affects farming activities with high water demand. In New York, the 2016 drought negatively impacted the local economy. Drought forecasting for New York State is much needed. Farmers in the region are walking for answers. The US Drought Monitoring is the most utilized tool, employing multiple indices and climatic variables to assess drought. However, the ability to forecast drought at the intra-seasonal scale (30-60 days) is underdeveloped.

At Cornell University, we are developing a forecasting product for New York State. Our initial version, while promising, has a coarse resolution that limits its utility at a farm scale. However, our ongoing research holds the potential to significantly impact the farming industry, as it aims to provide a higher-resolution forecasting product that can better assess the impact of drought on farming productivity.

Our recent publication has paved the way for a potential breakthrough in extending predictability for New York State. While our focus has been on the US Great Plains, our investigation into the sources of predictability using the low-level and upper levels wind has revealed a crucial teleconnection between the Great Plains and the Northeast, communicated via upper-level winds. This discovery has led us to incorporate this teleconnection mechanism into our second version, a development that we believe will significantly enhance drought predictability in New York.

Thus, the answer to moving forward on NYS drought predictability is in the winds. We hypothesize that upper- and lower-level winds keep the information about the complexity happening in the Great Plains and moving to NY. Currently, that information needs to be included in our old forecast system. Here, we test how low-level winds—that feed the inner continent moisture—rule the importance of drought predictability. Because we need a communication form that connects this remote process, we test what role the upper-level wind plays in this new configuration. We focus on summer because the wind effect on NY drought matures during that season when major agriculture occurs.

We used sophisticated multi-model datasets from the North American Multi-Model Experiment (NMME) to make this experiment possible. However, our expertise supports that it can work with any numerical modeling approach with the two wind components. In the end, we aim to use a Big Data/Machine Learning approach to generate the best drought forecasting product in NY that outperforms traditional approaches.