Disentangling the Role of Internal Variability in Atmospheric Vapor pressure Trend Biases Over the Southwestern United States

The vapor pressure deficit (VPD), or the difference between saturation vapor pressure (SVP) and actual vapor pressure (VP), is closely related to wildfire occurrence and spread in the Southwest United States (SWUS). Despite the importance of VPD, previous studies have found discrepancies between observations and model simulations of its VP component. To accurately simulate VPD, it is necessary to understand how internal variability and external forcing may affect observed VP trends.

In this study, we conduct a 30-year moving trend analysis for the SWUS vapor pressure related quantities using both observations and the NOAA Geophysical Fluid Dynamic Laboratory’s SPEAR simulations. Observed SVP exhibits a positive trend from 1940-2020 that can be attributed to external forcing using model simulations. Observed VP transitions from an increasing trend to a decreasing trend in the late 1970s - a pattern not reflected in the simulations, where VP increases over the historical period consistent with Clausius-Clapeyron scaling. This discrepancy from the observed VP trends indicates the presence of inherent model biases that prevent accurate representations of atmospheric moisture content. Moreover, the non-uniform observed VP trends warrant further examination into the role of internal variability as the source of the VP trend. Preliminary composite analysis of SWUS VP anomalies and sea surface temperature (SST) suggests that VP trends could be related to fluctuations in SST, further justifying a deeper delve into the influence of internal variability as a contributor to both the observed VP trends and SWUS VP model bias.