Underlying mechanisms of distinct west-to-east spatial patterns of soil moisture-precipitation feedbacks across the conterminous United States
OA Version
Citation
Abstract
It is critical to understand and predict changes in the patterns of rainfall events in a region due to climate change because precipitation is the central resource in freshwater availability, agriculture, energy generation, and the ecosystem. Thanks to enormous research efforts over the past few decades, we now know that land properties also play a role in understanding precipitation events, and that there are complicated feedbacks between the land and atmosphere called land-atmosphere (L-A) interactions. Soil moisture and precipitation are important components of these L-A feedbacks and thus have been studied in observations and models. However, there is still uncertainty and model-data disagreements in terms of the sign and magnitude of soil moisture-precipitation (SM-P) feedbacks. Thus, the community has focused on the understanding of atmospheric boundary layer (ABL) processes. The ABL is directly influenced by land surface properties and it transports fluxes of heat, moisture, and other scalars from the surface to the Free Atmosphere (FA).
This dissertation focuses on the interaction between land surface fluxes and boundary layer clouds to explain underlying physical mechanisms of distinct west-to-east spatial patterns of observed SM-P feedbacks across the Conterminous United States (CONUS) (Tuttle & Salvucci, 2016; 2017, hereafter the reference research). To this end, this dissertation documents three related research projects. In Chapter Two, we studied simulations of the ABL evolution with a simplified process-based cloud-topped boundary layer (Chemistry Land-surface Atmosphere Soil Slab, CLASS) model (Vilà-Guerau de Arellano et al., 2015). We modified the model to make it applicable to a wide range of environmental conditions (CLASS-L, Ryu & Salvucci, 2024). For the simulations, we used idealized environmental conditions representing over 11,000 cases and verified the CLASS-L model against Large Eddy Simulations (LES). In Chapter Three, we applied the CLASS-L model to explain the underlying mechanisms of observed SM-P feedbacks with the assumption that the predicted daily maximum cloud mass flux (M_DM) greater than a given criterion can represent precipitating clouds. We then conducted sensitivity tests of M_DM to changes in evaporative fraction (EF), which is proportioned to SM. We thus seek to explain observed SM-P feedbacks with a model linking EF to M_DM. The North American Regional Reanalysis (NARR, Mesinger et al., 2006) data sets in the summer months (June, July, and August) from 2002 to 2010 are the initial conditions to simulate the CLASS-L model. The research period overlaps with the reference research (Tuttle & Salvucci, 2016). In Chapter Four, we show that a simple regression model captures the underlying mechanisms explained in Chapter Three. The regression relates the M_DM to the early-morning environmental conditions and EF, and thus captures the totality of land-atmosphere interactions studied in this dissertation.
Based on the results of Chapter Three and Chapter Four, we explain the underlying mechanism of distinct west-to-east spatial patterns of observed SM-P feedbacks. The surface energy partitioning plays a critical role because the sensible heat flux is the source of the ABL growth. Thus, the depth of the ABL is deeper in the western CONUS and shallower in the eastern CONUS. On top of this fact, analyses of the drivers of cloud mass fluxes (a precursor to precipitation) reveal competing influences due to cloud core fraction, convective velocity, and humidity deficit, the net effect of which determines the strength and sign of the feedbacks. Furthermore, the CONUS-wide regression provides a more accessible tool than physical models to understand L-A interactions across CONUS. Beyond the processes evaluated in this dissertation, further knowledge of surface dynamics, cloud physics, interactions from the surface dynamics to clouds, and large-scale phenomena (such as convergence and divergence) are essential for a complete understanding of the totality of L-A interaction.
Description
2024
License
Attribution-NonCommercial 4.0 International