Contributions to Hydrometeorological Modeling: From Micro-Scale Land-Atmosphere Interactions to Large Scale Snowpack Impacts

Abstract

Hydrometeorological modeling (but in general geophysical modeling) is a valuable tool for scientists and stakeholders. It comprises several tasks, including (1) collection and curation of geophysical variables and parameters, to design and test the models, (2) formulation of a mathematical model, through the identification of relationships between such variables and parameters, and (3) coding, testing, and evaluation of the model. Furthermore, modelers often go one step further, analyzing model results to (4) further improve the models, and (5) to advance the physical understanding of the Earth System functioning. Here, a set of works addressing such tasks will be introduced.Arevalo et al. (2020a) addresses the aforementioned task 1, by organizing, quality controlling, making available, and describing a dense (in time and space) dataset of observations (about 70,000,000 single observations) within the Biosphere 2 - Landscape Evolution Observatory (LEO). This work provides high temporal resolution (i.e., 15 minutes) observations of (a) soil variables (e.g., soil moisture and temperature) near the surface (i.e., 5 cm depth) of the three near-identical artificial hillslopes (11x33 m2, ~1 m depth), (b) atmospheric variables (e.g., air temperature and humidity, wind speed) at 5 different heights (0.25 m to 9-10 m above ground) within the controlled enclosing atmosphere of LEO’s glasshouses, and (c) spatially aggregated variables (e.g., precipitation, relative water storage, and discharge). It is expected that sharing this data contributes to improve our understanding of the micro-scale land-atmosphere interactions, thus conducting to improved models. Tasks 2, 3, 4 and 5 are addressed in Arevalo et al. (2021), where daily snow water equivalent (SWE) from the UA-SWE dataset in conjunction with precipitation and temperature from PRISM data were analyzed. Based on the identified relationships, a simple daily SWE scheme was developed (tasks 2 and 3) to contribute to the improvement of the Climate Prediction Center (CPC/NCEP/NOAA) water balance model (LB), currently used operationally for drought monitoring and prediction (task 4). Evaluation of SWE simulated by the new scheme shows a good agreement with observations (UA-SWE and SNOTEL data), outperforming a complex Land Surface Model (LSM; i.e., Noah/NLDAS-2). The addition of snowpack treatment to LB (LBS, with S denoting snow) impacts the annual cycle of soil moisture leading to an improved representation of the drought condition with respect to the original LB (objective, based on standardized anomalies of LBS/LB simulated soil moisture) when compared to the U.S. Drought Monitor (U.S.-DM; subjective, expert multi-variable analysis), highlighting the importance of snowpack for drought assessment (task 5). LBS is currently operational globally providing a drought assessment that consider more key processes of the water cycle than the traditional drought indices, while requiring only data usually available. Work in Chapter 3 describes the analysis of 1 April SWE over CONUS and identify the ratio of cumulative ablation to accumulation (ABL/ACC) as an important source of information, explaining most of the 1 April SWE interannual variability (task 5). Furthermore, forest cover and elevation are found to highly modulate the strength of the linear relationships between 1 April SWE and its drivers, but only at some elevation bands. Additionally, sensitivity of 1 April SWE to temperature and precipitation is explored through linear regression models and by perturbing the forcing temperature and precipitation in the LBS model (task 5), showing that most of the mountainous West is highly vulnerable to the impacts of climate change. Finally, two works currently underway are described: (1) a continuation of the 1 April SWE analysis that explores the drivers of its trends, and other that pursues incremental improvements to a complex model as the operational implementation of WRF for Western U.S. utilized by the Center for Western Weather and Water Extremes (CW3E) through the improvement of the initial representation of the snowpack (task 4).