Corn Belt pressure systems analyzed using a manual synoptic classification method and self-organizing maps for distinguishing cropland and deciduous forest sensible and latent heat fluxes

Synoptic climatology has historically relied on manual classification of atmospheric pressure/height patterns, which are increasingly being replaced by "objective" methods, particularly self-organizing maps (SOMs). However, there have been few comparisons made between SOMs and manual synoptic classifications to determine which classification system best performs for addressing novel climatic research questions. Here we compare a manual synoptic classification of mean sea level pressure patterns over the U.S. Corn Belt to SOM derived synoptic classifications. We show that the SOM algorithm can successfully generate physically meaningful synoptic pressure classifications that are analogous to the manual classification. However, manual classifications may still need to be applied to detect transitional or rare synoptic states that the SOM algorithm may merge into broader classes. Nonetheless, the SOM-derived synoptic classification identifies synoptic type(s) favoring a low-pressure system to the west and a high-pressure system to the east of the Corn Belt that coincide with days having statically significant differences in latent and sensible heat fluxes between cropland and deciduous forest AmeriFlux towers. Our findings show that SOMs can successfully detect the synoptic types that maximize the impact of Corn Belt land use/land cover (LULC) types on heat fluxes, making them a useful tool in ongoing studies that seek to identify the synoptic conditions occurring when and where LULC is likely to trigger convective precipitation.