Embedded Files
Project Description
Background: Drought and soil moisture deficits exert important influence over ecological dynamics across the country. In addition, climate change is expected to alter soil moisture, leading to more severe deficits and enhanced aridity in many places.Despite recognition of these potentially important shifts, projections for future soil moisture patterns across space, through time, and within the soil profile are largely unavailable. Resource managers need soil moisture projections to make informed decisions about long-term conservation and preservation investments. There are currently very few ecologically appropriate, spatially broad datasets about future drought at resolutions suitable for informing natural resource decision making. This project will meet some of those needs by applying a water balance model that was designed to represent the impact of climate on dryland vegetation. Results will include detailed and synthesized drought information for the 21st century across the coterminous U.S. that are delivered in a format suitable for hosting on the Climate Toolbox, an established source for long-term climate projections. Data provided by this project will be useful for a wide variety of applications, including scenario planning, climate adaptation frameworks, species distribution models, and habitat vulnerability assessments.
Objective: Developing quantitative projections of future patterns of soil moisture and ecological drought that determine ecosystem vulnerability to drought under climate change is essential for prioritizing management actions and determining the most effective adaptive strategies. Despite the importance of detailed soil moisture patterns for understanding and managing ecosystems, high-resolution, ecologically relevant measures of drought are not readily available. This project will help meet those needs by developing and disseminating high resolution (~4 km) gridded, 21st century projections of soil moisture in the MACA-gridMET framework across the coterminous United States. Although we will generate results for this entire area, our measures of soil moisture and ecosystem water balance are most appropriate for the dryland regions and ecosystems.
Approach: We identified five essential characteristics of drought and water balance approaches that are necessary to appropriately assess future dryland ecological drought in the context of global change:
1. Frequent temporal resolution: Water balance models must employ a reasonably frequent temporal resolution to capture ecologically-relevant drought conditions that can emerge rapidly; fluctuating dry soil events are often not well-represented by meteorological drought indices (Walker et al 2023).
2. Site-specific soil conditions: Models must represent the influence of site-specific soil conditions, including the soil moisture release curve, to appropriately estimate water balance, water movement and the availability of that water for utilization by plants (Novick et al. 2022).
3. High resolution across soil profile: Models need to sufficiently resolve the soil profile at multiple depths to represent differential water movement and utilization among plant functional types and across seasons. Seasonal and soil-depth patterns of moisture availability strongly influence recruitment as well as the structure and function of dryland plant communities (Romme et al. 2009; Samuels-Crow et al. 2020).
4. Vegetation structure and physiology: Models must incorporate the influence of vegetation, which include vegetation structural impacts on the fate of incoming precipitation, hydraulic redistribution and vegetation physiology that determines patterns and rates of plant water use (Andrews et al. 2020).
5. Climate change impacts on vegetation: Accurately estimating future ecological drought requires a water balance model that includes the consequences of climate change on vegetation structure (Tietjen et al. 2017) and function including responses to increased atmospheric carbon dioxide (Köhler et al. 2015). Meteorological drought indices and strictly physical water balance models will overestimate the ecological severity of future drought conditions and provide misleading impressions ecological impacts (McColl et al. 2022).
We are utilizing SOILWAT2 (Schlaepfer et al. 2023), an ecosystem water balance model that meets the requirements above. SOILWAT2 is a process-based daily simulation model that represents the soil profile with multiple soil layers and allows the surface to be comprised of a snowpack, unvegetated ground, and vegetation composed of multiple co-occurring plant types responsive to atmospheric CO2 concentrations. The code is available as open-source R package rSOILWAT2. SOILWAT2 represents evaporative demand accounting for local topography, interception by vegetation and litter, evaporation of intercepted water, snowpack accumulation and ablation, snowmelt, infiltration, soil temperature, saturated and unsaturated percolation and hydraulic redistribution for each soil layer, bare-soil evaporation, transpiration, and deep drainage. The model has successfully been applied to North American and global dryland ecosystems.
Simplified conceptual diagram showing water fluxes (arrows) and pools (ovals) in SOILWAT2.
Progress
We have made progress on this multi-year project over the last year!
Model Development
The goal was to increase input, output, and throughput capabilities of the simulation program. Specifically, we have implemented five main areas of improvements
Representation of the influence of soil organic matter on the soil water retention curve and the saturated hydraulic conductivity parameter
Calculation and output of ecological drought metrics during the simulation runs directly by SOILWAT2
Writing outputs as netCDFs following CF conventions using the netcdf-c library (SOILWAT2 v8.0.0)
Handling inputs from a variety of netCDF data sources including unit conversions using the udunits2 library (SOILWAT2 v8.1.0)
HPC-ready parallelization using the OpenMPI library (release coming soon)
Model testing
We collaborate with Tyson Ochsner, Erik Krueger, Zach Hoylman and others on a project to validate observed soil moisture at c. 1,000 stations across the conterminous U.S. with soil moisture simulation models. First, preliminary results are just coming in and they look promising even for SOILWAT2 runs without the new vegetation representation!
This map shows the agreement between daily simulated soil moisture and observations at c. 1,000 stations as measured by the Kling-Gupta efficiency (KGE) which combines metrics of correlation, variability of prediction errors and bias. The prelimary results suggest a decent agreement between simulated SOILWAT2 soil moisture in magnitude and day-to-day variability.
This figure summarizes correlations between daily simulated soil moisture and observations at c. 1,000 stations for the total soil profile and three soil depths separated by six station networks. The prelimary results suggest overall decent agreement and indicate areas for future model improvements (e.g., SNTL stations). The low agreement with NEON stations may indicate also areas for model improvements and/or pending issues with NEON data processing.
CONUS-wide representation of vegetation under contemporary and future climates
coming soon...
Collaborators
Imtiaz Rangwala - University of Colorado at Boulder
Katherine Hegewisch - University of California, Merced
John Gross & David Thoma - National Park Service
John Guinotte - US Fish and Wildlife Service
Support
USGS Climate Adaptation Science Centers
Andrews, C. M. et al. Low stand density moderates growth declines during hot droughts in semi‐arid forests. J Appl Ecol 57, 1089–1102 (2020).
Chambers, J. C. et al. New indicators of ecological resilience and invasion resistance to support prioritization and management in the sagebrush biome, United States. Front Ecol Evol 10, 1–17 (2023).
Chenoweth, D. A. et al. Ecologically relevant moisture and temperature metrics for assessing dryland ecosystem dynamics. Ecohydrology 16, e2509 (2023).
Köhler, I. H. et al. Last-century increases in intrinsic water-use efficiency of grassland communities have occurred over a wide range of vegetation composition, nutrient inputs and soil pH. Plant Physiol 170, 881–90 (2015).
McColl, K. A. et al. The terrestrial water cycle in a warming world. Nat Clim Chang 12, 604–606 (2022).
Novick, K. A. et al. Confronting the water potential information gap. Nat Geosci 15, 158–164 (2022).
Romme, W. H. et al. Historical and Modern Disturbance Regimes, Stand Structures, and Landscape Dynamics in Piñon–Juniper Vegetation of the Western United States. Rangeland Ecol Manag 62, 203–222 (2009).
Samuels-Crow, K. E. et al. Atmosphere-Soil Interactions Govern Ecosystem Flux Sensitivity to Environmental Conditions in Semiarid Woody Ecosystems Over Varying Timescales. J Geophys Res-Biogeo 125, e2019JG005554 (2020).
Schlaepfer, D. R. et al. SOILWAT2: An Ecohydrological Ecosystem-Scale Water Balance Simulation Model. Version 7.0.0. (2023) https://github.com/DrylandEcology/SOILWAT2
Tietjen, B. et al. Climate change-induced vegetation shifts lead to more ecological droughts despite projected rainfall increases in many global temperate drylands. Glob Change Biol 23, 2743–2754 (2017).
Walker, D. W. & Van Loon, A. F. Droughts are coming on faster. Science 380, 130–132 (2023).
Page updated
Google Sites
Report abuse