Semi-arid and arid environments are characterized by low precipitation and high evapotranspiration (ET), leaving little water available for discharge into surface water bodies and groundwater recharge. For these water-limited environments, understanding the relationships between precipitation, ET, and soil moisture is critical. These relationships not only affect water resources in these increasingly populated regions but are also necessary to predict the impact of climate change on semi-arid and arid ecosystems.
The overall goal of this dissertation was to shed light on the quantitative relationships between precipitation, evaporation, ET and soil moisture dynamics in an arid and a semi-arid environment. A three-step approach was chosen to (i) quantify soil water fluxes and storage of bare, arid soil in weighing lysimeters, (ii) evaluate a process-based evaporation model using evaporation data from the same bare, arid soil, and (iii) monitor the soil moisture and temperature dynamics of a vegetated, semi-arid soil at a field site. The studies were carried out at two sites in Nevada: (1) The bare arid soil is part of the SEPHAS weighing lysimeter facility, located in Boulder City near Las Vegas, southern NV; (2) The vegetated semi-arid soil with a shallow water table (depth to water
In a first step, precipitation, evaporation and infiltration, total soil mass and soil moisture profiles were monitored in three lysimeters over four years (water years 2009-2012). The results showed that 88 out of 180 precipitation events occurred during winter, but the average amount of precipitation per event was the highest during summer. Between 69% and 90% of annual precipitation was found to evaporate back into the atmosphere during the course of a water year (October through September). Water years with large amounts of winter precipitation (water years 2010 and 2011) yielded higher water storage compared to winters with lower amounts of precipitation (water years 2009 and 2012). Throughout spring, summer and autumn, most of the precipitated water evaporated back into the atmosphere, even after high intensity storms.
As for the soil moisture profile, two precipitation thresholds were found. The first threshold ranged between 0.5 and 2 mm total precipitation, representing the smallest amount of precipitation to cause a change in soil moisture content > 0.01 m3 m-3 at 2.4 cm depth. This range depended on season, antecedent conditions, and the amount of time between previous and successive events. Data show that events with less than 1-2 mm total precipitation have little to no impact on moisture content below the immediate soil surface (top inch or so). With respect to soil water storage, water from storms with 1-2 mm total precipitation evaporate within a day, and do not have an impact on long-term soil water storage. The second precipitation threshold could be defined as the smallest amount of total precipitation needed to change soil moisture content at 25 cm depth by > 0.01 m3 m-3. Only 14 out of the 180 events (or sequence of events) changed soil moisture content at 25 cm depth or deeper. The 10 largest precipitation events (with respect to total amount of precipitation) were analyzed in more detail. Following all 10 events, soil moisture at 25 cm increased by 0.01 m3 m-3 or more with total precipitation, intensity, and duration ranging between 13.2-41.6 mm, 0.6-12.3 mm hr-1, and 1.5-52 hours, respectively. In addition, there were two additional events and periods of time where multiple events occurred in one or two days that also increased soil moisture at 25 cm by more than 0.01 m3 m-3 (precipitation was 10.4 mm or greater). During the study period, only 7 events (or sequences of events) changed soil moisture content down to 50 cm depth. It took just under 4 years to see an increase in soil moisture at a depth of 250 cm. The moisture profiles in the lysimeters indicate that most of the moisture dynamics (infiltration as well as evaporation) occurs within the top 25 cm. Precipitation that infiltrates below 25 cm seem to remain in the soil and fosters further infiltration during and after storm events while evaporating only slowly in between events.
For the second step, a recently developed process-based evaporation model by Shokri et al. (2009) and Or et al. (2013) was employed to simulate measured evaporation rates from the weighing lysimeter soil already discussed above. The model focusses on water-vapor diffusion-controlled (or Stage II) evaporation and calculates Stage II evaporation rates based on soil texture, total porosity and an initial moisture content profile as input parameters. Simulations of the evaporation rates using readily available soil physical properties agreed well with two out of the three events that were analyzed (evaporation rate RMSEs of 0.093 and 0.141 mm d-1, respectively). For the third event, simulations systematically underestimated measured evaporation rates (RMSE of 0.181 mm d-1). The latter was likely due to considerable differences between the moisture profile in the lysimeter soil compared to the simplified moisture profile that is assumed by the model. Monte-Carlo simulations showed that total porosity and difference in soil moisture content above and below the second drying front are the models most sensitive parameters. Since total porosity can be determined rather accurately, improving the soil moisture profile characterization would likely need more scrutiny to further improve the model predictions for arid soils.
In a third step, moisture and temperature dynamics were monitored for a vegetated, semi-arid soil at a field site in the Great Basin over one year (July 2010 to June 2011). The question was whether any surface water from precipitation or snow melt would percolate all the way to the groundwater table located at five to six meters below the soil surface and would therefore provide groundwater recharge. The soil consisted of a loam / sandy clay loam overlying clayey material containing lenses of sand and silt and was mostly vegetated by phreatephytes and other shrubs. Inter-annual and seasonal precipitation, ET, soil moisture and temperature were determined. With Fiber Optic Distributed Temperature Sensing (FO DTS), a novel technique was employed to measure temperature profiles in the soil at ~1 cm spatial resolution. The results show that moisture from precipitation and snow melt percolates as deep as ~400 cm but does not reach the phreatic zone. Although, the TDR probes were placed in the soil vertically with the addition of backfill different than the surrounding soil, potentially creating a conduit for water flow. Changes in soil moisture content were observed at 500 cm depth but likely due to changes in groundwater table rather than percolating water from the soil above. Soil moisture and ET data show that infiltration leads to soil moisture gain from October to December whereas moisture loss due to ET dominates from March through September. FO DTS results showed diurnal variation of soil temperatures down to ~50 cm depth and seasonal variations down to ~500 cm were observed. Sensors recorded multiple cold wetting fronts in March through April 2011. Changes in soil temperatures were observed that related to changes in soil water storage; however, this occurred only near surface.
In conclusion, this dissertation sheds some light on how arid and semi-arid soils infiltrate, store and evaporate water as functions of precipitation, atmospheric demand and antecedent soil moisture conditions. The studies described above provide a starting point with respect to the hydraulic behavior of desert soils and a first set of baseline data. The SEPHAS weighing lysimeter has been in operation since 2008 and there are currently seven years of soil, mass, moisture content, matric potential and temperature data. To date, four years of data analysis have been completed but the methods in place can now easily be applied to the entire seven years. Overall, increased knowledge about the role of desert soils in the hydrologic cycle will serve as well in the long run, in particular with the ever increasing pressure on water resources globally, the anticipated shifts in precipitation patterns due to climate change as well as the increasingly frequent and longer droughts.
Data sets used in this project were unusually large, and are reported in their entirety as supplementary materials as Appendices A and B. In Appendix A, supplementary data include multiple files from the SEPHAS Weighing Lysimeter Facility. Files are organized into folders by Lysimeter (1-3) or Other (meteorological). Data files in Lysimeter 1-3 folders are named by lysimeter number, instrument, hydrologic year, and data type. For example, theta data (also known as water content data using TDR probes) from Lysimeter 1 during the hydrologic year 2008-2009 would be in the Excel file titled “Lysimeter 1_TDR_Hydro Year_2008-2009_Theta.xlxs”. The one data file in the Other folder titled “Meteorological Station_2008-2012_Meteorological Data.xlxs” includes four years of meteorological data (wind speed, air temperature, vapor density, relative humidity, precipitation, net radiation, atmospheric pressure, and air density). Supplementary data in Appendix B includes five Excel spreadsheets for the Spring Valley Site SV6. These files are titled via site, instrument, time period, and data type. For example, water content data using TDR probes from 2010-2011 would be in the Excel file titled “SV6_TDR_2010-2011_Water Content.xlxs”.