Skip to main content
ARS Home » Pacific West Area » Tucson, Arizona » SWRC » Research » Research Project #432380

Research Project: Understanding Water-Driven Ecohydrologic and Erosion Processes in the Semiarid Southwest to Improve Watershed Management

Location: Southwest Watershed Research Center

2022 Annual Report

1:As part of the LTAR network, and in concert with similar long-term, land-based research infrastructure in the region, use the Walnut Gulch LTAR site to improve the observational capabilities and data accessibility of the LTAR network and support research to sustain or enhance agricultural production and environmental quality in agroecosystems characteristic of the semiarid Southwest region. Research and data collection are planned and implemented based on the LTAR site application and in accordance with the responsibilities outlined in the LTAR Shared Research Strategy, a living document that serves as a roadmap for LTAR implementation. Participation in the LTAR network includes research and data management in support of the ARS GRACEnet and/or Livestock GRACEnet projects. 1.1:Improve & continue long-term measurements & analysis of water budgets on WGEW & Santa Rita Experimental Range (SRER) watersheds. 1.2:Expand variables measured on WGEW & SRER watersheds based on recommendations of the LTAR Meteorology, Hydrology, CO2, Non-CO2 Gas, Soil, Biology, & Wind Erosion Committees. 1.3:Develop a long-term monitoring program. 1.4:Implement an experiment on the SRER watersheds to quantify the effects of brush management on a set of ecosystem services. 1.5:Compute trends in sub-daily & daily precipitation intensity across LTAR sites. 1.6:Evaluate National Weather Service dual pole radar precipitation data & its ability to improve flash flood forecasting. 2:Quantify how seasonal, annual, and decadal-scale variations in climate, plant community composition, and management impact processes controlling the cycling of water, energy, and carbon in semiarid rangelands 2.1:Determine how changes in vegetation structure & climate affect ecosystem-atmosphere water vapor & CO2 exchange using long-term flux tower observations. 2.2:Use isotopes in pond deposition sediments to understand & quantify erosion & sediment yields in semiarid landscapes as a function of ecological sites. 2.3:Quantify the impact of erosion control structures on runoff & sediment transfers in semiarid landscapes. 2.4:Estimate annual production & minimum total foliar cover using Landsat & MODIS satellite. 2.5:Develop methods to assess climate impacts on rangeland vegetation composition & production across the West. 3:Develop a new conceptual framework and corresponding experimental methods to understand and model the dynamics of semiarid upland and channel erosion processes. 3.1:Conduct experiments to quantify the effects of surface condition. 3.2:Conduct experiments to develop a remote sensing method to estimate hydraulic roughness. 4:Improve hillslope (RHEM) and AGWA/KINEROS2 watershed models and develop methods to incorporate new remotely sensed, meteorologic, & land surface information. 4.1:Complete development & post-disturbance testing of the RHEM for application in Western rangelands. 4.2:Develop a mechanism to extend the findings from the Walnut Gulch LTAR site across Arizona & New Mexico & support collaborative vegetation management of public lands to improve watershed function. 4.3:Incorporate a variety of KINEROS2 (K2) / AGWA model enhancements.

Objective 1: 1. Use co-located rain gauges to quantify uncertainties in long-term precipitation datasets. 2. Use radar stage measurements to test remote methods to measure runoff stage 3. Deploy mobile x-band Doppler radar and compare with Dual Pole radar rainfall rain-gauge observations, and runoff observations on the WGEW. 4. Meet LTAR objectives by: a) using observational datasets to quantify the individual components of the watershed water balance in Walnut Gulch Experimental Watershed WGEW), b) using satellite and ground measurements of vegetation to document changes in watershed vegetation, c) determining trends and magnitude of precipitation intensities and precipitation extremes across the continental US, and d) implementing the LTAR common experiment to assess the effects of brush management on a set of ecosystem services. Objective 2: 1. Use long-term flux tower observations to determine how changes in vegetation structure and climate affect ecosystem-atmosphere water vapor and carbon dioxide exchange. 2. Use 210Pb pond stratigraphy to determine erosion rates and their historical dynamics on small watersheds over the past 50-100 years. 3. Quantify runoff and sediment yields on watersheds to quantify the impact of erosion control structures on runoff and sediment transfers. 4. Use satellite, climate, site productivity and management data to estimate annual production and minimum total foliar cover. 5. Use LiDAR, point cloud, and new satellite datasets to construct canopy height models to assess climate impacts on rangeland vegetation composition and production. Objective 3: 1. Use rainfall simulator experiments to quantify the effects of surface condition on infiltration, runoff, concentrated flow dynamics, sediment transport processes, and surface evolution. 2. Use radar backscatter roughness and hydraulic roughness at a laboratory, rainfall simulator, and small watershed scales using airborne and satellite active radar imagery to develop a remote sensing methods to estimate hydraulic roughness. Objective 4: 1. Complete development and post-disturbance testing of the Rangeland Hydrology and Erosion Model (RHEM) for application in Western rangelands. 2. Create a web interface to identify problem areas in watersheds, compare across watersheds, and assess trends in time prior to KINEROS2 modeling. 3. Incorporate RHEM, improved process model representations, and higher-resolution, model inputs, sub-surface and variable width routing, and interstorm processes into KINEROS2.

Progress Report
This is the final report for project 2022-13610-012-000D, which has been replaced by new project 2022-13610-013-000D, "Understanding Ecological, Hydrological, and Erosion Processes in the Semiarid Southwest to Improve Watershed Management." For additional information, see the new project report. For Sub-objective 1.1.1, rainfall was measured with above ground and pit rain gauges. Rainfall undercatch was observed at three locations within the Walnut Gulch Experimental Watershed (WGEW). These observations were collated with nearby wind speed and direction data as well as with video and one-dimensional disdrometers. Complete analysis was delayed due to a vacancy. 1.1.2: A down-looking radar and a down-looking radar coupled with an up channel-looking radar were deployed at flumes 6 and 4 in WGEW to sense runoff velocity. Radar measurements at Flume 6 were within 10% of stilling well stage measurements, but they were not at Flume 4. We worked with our collaborators to reduce the sampling interval and will re-evaluate the Flume 4 sensor during the next project period. 1.1.3: Complete water budgets quantifying annual precipitation, runoff, change in surface soil moisture, and evapotranspiration were developed for watersheds in WGEW. The water balance was closed (i.e., inputs = outputs) within 5-10% annually. These results were used to assess the accuracy of eddy covariance measurements of evapotranspiration and in a comparison of water budgets across many Long-Term Agroecosystem Research (LTAR) locations. 1.2: A recommended cross-LTAR wind-shielded rain gauge has been deployed in Tombstone for five years adjacent to a standard weighing rain gauge WGEW and disdrometers for comparison. Real-time data access based on the Campbell cloud database and dashboards was tested. 1.3: A network of 18 vegetation monitoring sites across WGEW was established, along with monitoring protocols to quantify canopy and ground cover, basal gap, plant height and species. All sites have been monitored at least once. Imagery has been collected for concurrent years via drone. The datasets have been used to quantify brush cover using Landsat imagery for the Rangeland Brush Estimation Tool (RaBET). 1.4: A herbicide brush management treatment was applied to a mesquite-invaded rangeland in June 2016. The treatment was unsuccessful in killing the mesquite trees. Mesquite foliar cover rebounded from 10% in July 2016 to 70% in October 2019. Herbaceous diversity and biomass under mesquite canopies increased following treatment. Below-average precipitation in 2017 caused decline in biomass on control and treated areas. Pre- and post-treatment levels of soil organic C and total N was assessed. Data on small mammals and arthropods were collected by collaborating biologists. 1.5: Trends in sub-daily rainfall intensities were investigated across the LTAR watershed network Daily Cooperative Observer (COOP) network rainfall observations were compared to nearby ARS daily extreme observations. The LTAR network of precipitation gauges was used as an extra layer of quality control for the COOP network. Results showed consistency in precipitation extremes between the COOP and LTAR networks. Despite discrepancies at daily time steps, the extreme precipitation observed by COOP rain gauges can be reliably used to characterize changes in the hydrologic cycle due to natural and human causes at daily time scales. Next, precipitation intensification data at sub-daily time scales available since the 1970s from the ARS network were examined. These data show that summer monsoon rainfall has intensified at WGEW. However, the spatial extent of monsoon convective storms has not changed with time. Comparison across ARS LTAR locations was not completed due to inconsistencies in trends using multiple methods. 1.6: A mobile radar was deployed in WGEW during the summer of 2017 and 2018. The operating collaborator was unable to process its X-band radar due to lack of staff. Comparison between WGEW rain observations and National Weather Service radar products: digital hybrid reflectivity (DHR), digital precipitation rate (DPR), and multi-radar/multi-sensor (MRMS) precipitation rates were completed. DPR provides better precipitation estimates than DHR. DPR suffered from mountain blockage of the radar signal that was improved by the MRMS product. For Sub-objective 2.1.1, studies were conducted in a riparian woodland site to understand the trajectory and magnitude of riparian vegetation water use. These studies highlighted the role of hydraulic redistribution by mesquite roots in different physiographic settings in vegetation water use and carbon uptake. New techniques to partition total evapotranspiration (ET) into plant transpiration and abiotic evaporation were applied to better understand the interannual variability of riparian water use. For 2.1.2, studies were conducted to understand how climate-driven changes in water availability may affect net carbon dioxide exchange across a broad diversity of ecosystems and climates, from rangelands to mountain forests, in the southwestern United States. 2.1.3: An analysis of how ecosystem leaf phenology interacts with soil water availability was conducted to better understand how shrub springtime productivity versus summer grass productivity contributes to annual ecosystem productivity. 2.1.4: A paired flux tower comparison quantified how brush management affects mesquite-encroached grassland land-atmosphere water and carbon fluxes. The mesquite treatment had only short-lived effects on fluxes, but this was likely due to the failure of the treatment to permanently kill the mesquite. Despite the lack of treatment efficacy, persistent differences between the experiment and control site were found, which was likely due to differences in vegetation composition and soil type. 2.2: Isotopes were collected in stratified sediment accumulated in stock ponds to quantify sediment yields from rangelands. The experiments point to the use of isotope analysis to quantify changes in rangeland soil erosion and sedimentation impacted by conservation structures. 2.3: Studies were conducted to quantify sediment accumulation and watershed outlet sediment yields in watershed treated with check dams. Although check dams reduce sediment yield initially, the reductions are not persistent. Field experiments are ongoing to analyze isotopes in sediment accumulated behind check dams to quantify erosion and sedimentation rates. Sub-objective 2.4 was abandoned as several west-wide biomass and cover products have been developed by extending Natural Resources Conservation Service (NRCS) and Bureau of Land Management datasets with Landsat data. Instead, machine learning models were built to estimate biomass and cover predicted using national methods. Sub-objective 2.5 was abandoned as the impact of climate on vegetation were incorporated through reviews of the Climate Hub-led Grass-Cast tool in Arizona with collaborators. For Sub-objective 3.1, rainfall simulator experiments were conducted on plot-scale soil boxes to quantify the effects of surface condition on infiltration, runoff, flow dynamics, sediment transport, and surface evolution. Surface roughness and erosion change over time were quantified under simulated rainfall conditions. Steeper slopes become rougher than shallower slopes although erosion was approximately the same. These results are important for understanding how surface soil conditions affect erosion. Sub-objective 3.2 was not completed because our university partner was unable develop a laboratory radar unit for the proposed project. For Sub-objective 4.1, the Rangeland Hydrology and Erosion Model (RHEM) hillslope engine was incorporated into KINEROS2 (K2) and Automated Geospatial Watershed Assessment (AGWA). Thus, K2 has two options for hillslope infiltration and erosion. New tutorials were developed for web-based RHEM and for K2-RHEM, and RHEM version V2 was released with updated documentation. Time varying remotely-sensed land cover data was incorporated into K2/RHEM over a 20-year period for a small sub-watershed of the WGEW. The long-term average of remotely sensed vegetation cover from 1996-2014 provided the best estimate of simulated runoff when compared to measured data. The framework provides a means for a simple, parameterization for watershed-scale modeling where ground-based vegetative data are limited. Evaluation of post-fire modeling was conducted based on burned areas within the LTAR ARS Reynolds Creek Experimental Watershed., but no post-fire runoff and erosion events were recorded for several years during which the watershed recovered to near pre-fire conditions. Post-fire watersheds that did have runoff-erosion events were used to evaluate post-fire K2-AGWA parameters and improve post-fire look-up table parameters. 4.2: Spatial databases of land ownership were created, and reports from publicly available remote sensing information for specific land units were generated, but the pandemic prevented input from watershed groups and land managers. 4.3: The K2/AGWA model was improved to: 1) Derive representative 2-D hillslope ground surface profiles from 3-D digital data; 2) Represent the impacts of military training activities on rangelands; 3) Couple AGWA and the facilitator decision support tool to enable use with a multi-objective decision making framework that compares choices in a structured manner; 4) Ingest radar-rainfall estimates from new NWS dual-pole radars as well as near-real time, telemetered rain gauge intensity measurements to improve flash flood forecasting including testing at several NWS Weather Forecast Offices; 5) Incorporate high-resolution LIDAR derived topographic data; 6) route flow using a variable hillslope width; 7) Estimate flood inundation from peak discharge estimates; and, 8) Enable execution on high performance computers.

1. Low-cost, flexible monitoring system helps understand Colorado River snowpack. Snowpack monitoring in forested settings is uneven across the west, with monitoring traditionally done in accessible locations and in level clearings where snowpacks are not influenced by trees or local topography. In reality, elevation, vegetation, slope, aspect, and other variables strongly control accumulation or loss of snow and the spatial distribution of snow and soil moisture. Even within very short distances, differences in key processes regulating snow accumulation and loss may result in dramatically different amounts and timing of snowmelt water. In collaboration with academic partners and The Nature Conservancy, ARS researchers in Tucson, Arizona, wrote and distributed a practitioner’s handbook of Snowtography, a low-cost, flexible system of snowpack monitoring based on automated remote cameras. This manual provides guidance on site selection, equipment, budgeting, fabrication, installation, and operation of a Snowtography station. Optional upgrades including remote data access and soil moisture monitoring. The handbook has been viewed over 1000 times and downloaded over 200 times.

2. Precipitation variability impacts carbon sequestration in semiarid grasslands. Precipitation amount is a strong control of carbon uptake and biomass production in grasslands, but little is understood about the impacts of temporal changes in precipitation, such as receiving the same annual rainfall amount packaged in fewer, larger storms. In collaboration with Chinese scientists, ARS researchers in Tucson, Arizona, conducted a three-year field experiment on plots in a semiarid grassland of Inner Mongolia in which they applied the same amount of precipitation repackaged into many small events or few large events, and they measured the carbon uptake and productivity. They compared these results to those of an ecosystem model, which is a key tool in predicting the impacts of future precipitation variability on agroecosystem carbon cycling. They found that most metrics of grassland productivity were maximized at intermediate levels of precipitation variability. In contrast, the model incorrectly predicted a linear increase in productivity across increasingly frequent, but also increasingly small, precipitation events. These results indicate that while agroecosystem models may correctly respond to changes in annual precipitation amounts, further work is needed to model the consequence of an increasingly variable hydrologic cycle.

3. Increasing rainfall intensities pose erosion risk for cultural resources. Projected increases in high intensity rainfall in the southwestern United States pose major threats to culturally important historical adobe structures managed by the National Park Service (NPS). In collaboration with the NPS Vanishing Treasures Program, ARS scientists from Tucson, Arizona, utilized ARS rainfall simulation technologies to evaluate the effects of varying rainfall intensity and amounts on erosion of traditionally built earthen adobe architecture. Surface erosion of adobe structures increased significantly with increasing rainfall intensity. The study results suggest that more frequent high intensity rainfall events will cause increasing damage to historical adobe structures throughout the southwestern United States. The findings provide critical advancements in understanding the impacts of a changing climate on important cultural resources and in framing preservation and restoration strategies for the NPS.

Review Publications
Goodrich, D.C., Bosch, D.D., Bryant, R.B., Cosh, M.H., Endale, D.M., Veith, T.L., Kleinman, P.J., Langendoen, E.J., McCarty, G.W., Pierson Jr., F.B., Schomberg, H.H., Smith, D.R., Starks, P.J., Strickland, T.C., Tsegaye, T.D., Awada, T., Swain, H., Derner, J.D., Bestelmeyer, B.T., Schmer, M.R., Baker, J.M., Carlson, B.R., Huggins, D.R., Archer, D.W., Armendariz, G.A. 2022. Long term agroecosystem research experimental watershed network. Hydrological Processes. 36(3). Article e14534. [Corrigendum: Hydrological Processes: 2022, 36(6), Article e14609.]
Li, L., Hao, Y., Zheng, Z., Wang, W., Biederman, J.A., Wang, Y., Wen, F., Qian, R., Xu, C., Zhang, B., Song, X., Cui, X., Xu, Z. 2021. Heavy rainfall in peak growing season had larger effects on soil nitrogen flux and pool than in the late season in a semiarid grassland. Agriculture, Ecosystems and Environment. 326. Article 107785.
Li, L., Hao, Y., Wang, W., Biederman, J.A., Wang, Y., Zheng, Z., Wen, F., Qian, R., Zhang, B., Song, X., Cui, X., Xu, Z. 2022. Joint control by soil moisture, functional genes and substrates on response of N2O flux to climate extremes in a semiarid grassland. Agricultural and Forest Meteorology. 316. Article 108854.
Barnes, M., Farella, M., Scott, R.L., Moore, D., Ponce-Campos, G., Biederman, J.A., MacBean, N., Litvak, M., Breshears, D. 2021. Improved dryland carbon flux predictions with explicit consideration of water-carbon coupling. Communications Earth & Environment. 2. Article 248.
Vivoni, E., Perez-Ruiz, E., Scott, R.L., Naito, A., Archer, S., Biederman, J.A., Templeton, N. 2021. A micrometeorological flux perspective on brush management in a shrub-encroached Sonoran Desert grassland. Agricultural and Forest Meteorology. 313. Article 108763.
Mahmud, K., Scott, R.L., Biederman, J.A., Litvak, M., Kolb, T., Meyers, T., Krishnan, P., Bastrikov, V., MacBean, N. 2021. Optimizing carbon cycle parameters drastically improves terrestrial biosphere model underestimates of dryland mean net CO2 flux and its inter-annual variability. Journal of Geophysical Research-Biogeosciences. 126(10). Article e2021JG006400.
Belmonte, A., Sankey, T., Biederman, J.A., Bradford, J., Kolb, T. 2022. Soil moisture response to seasonal drought conditions and post-thinning forest structure. Ecohydrology. 15(5). Article e2406.
Li, L., Qian, R., Liu, W., Wang, W., Biederman, J.A., Zhang, B., Kang, X., Wen, F., Ran, Q., Zheng, Z., Xu, C., Che, R., Xu, Z., Cui, X., Hao, Y., Wang, Y. 2022. Drought timing influences the sensitivity of a semiarid grassland to drought. Geoderma. 412. Article 115714.
Hart, S., Raymond, K., Williams, C.J., Johnson, J.C., Degayner, J., Guebard, M. 2021. Precipitation impacts on earthen architecture for better implementation of cultural resource management in the US Southwest. Heritage Science. 9. Article 143.
Zhang, F., Biederman, J.A., Pierce, N., Potts, D., Devine, C., Hao, Y., Smith, W. 2021. Precipitation temporal repackaging into fewer, larger storms delayed seasonal timing of peak photosynthesis in a semi-arid grassland. Functional Ecology. 36(3):646-658.
Yazbeck, T., Bohrer, G., Gentine, P., Ye, L., Arriga, N., Bernhofer, C., Blanken, P., Desai, A., Durden, D., Knohl, A., Kowalska, N., Metzger, S., Molder, M., Noormets, A., Novick, K., Scott, R.L., Siguit, L., Soudani, K., Ueyama, M., Varlagin, A. 2021. Site characteristics mediate the relationship between forest productivity and satellite measured solar induced fluorescence. Frontiers in Forests and Global Change. 4. Article 695269.
Meles, M.B., Demaria, E., Heilman, P., Goodrich, D.C., Kautz, M.A., Armendariz, G.A., Unkrich, C.L., Wei, H., Thiyagaraja Perumal, A. 2022. Curating 62 years of Walnut Gulch Experimental Watershed data: Improving the quality of long-term rainfall and runoff datasets. Water. 14(14). Article 2198.
Castellanos, A., Hinojo-Hinojo, C., Rodriguez, J., Romo-Leon, J., Wilcox, B., Biederman, J.A., Penuelas, J. 2022. Plant functional diversity influences water and carbon fluxes and their use efficiencies in native and disturbed dryland ecosystems. Ecohydrology. 15(5). Article e2415.
Young, A., Friedl, M., Seyednasrollah, B., Beamesderfer, E., Carrillo, C., Xiaolu, L., Moon, M., Arain, M., Baldocchi, D., Blanken, P., Bohrer, G., Burns, S., Chu, H., Desai, A., Griffis, T., Hollinger, D., Litvak, M., Novick, K., Scott, R.L., Suyker, A., Verfaillie, J., Wood, J., Richardson, A. 2021. Seasonality in aerodynamic resistance across a range of North American ecosystems. Agricultural and Forest Meteorology. 310. Article 108613.
Broadbent, C., Brookshire, D., Goodrich, D.C., Dixon, M., Brand, A., Thacher, J. 2022. Developing ecological endpoints for valuation of semi-arid riparian ecosystem services. Journal of Environmental Planning and Management.