Award Date


Degree Type


Degree Name

Doctor of Philosophy (PhD)



First Committee Member

Zhongbo Yu

Second Committee Member

Michael H. Young

Number of Pages



In arid and semi-arid regions such as in the southwestern United States, soil moisture is an essential component of desert ecosystems. Gaining better knowledge of moisture dynamics through appropriate numerical modeling will help us understand physical mechanisms that influence soil hydrologic processes in these regions. Moreover, numerical modeling of these processes is often emphasized because most desert watersheds are ungauged, and thus field observations are either not readily available or difficult to simulate. In this dissertation, three modeling studies were conducted to investigate the temporal and spatial soil moisture variation and hydraulic properties, and their effect on rainfall-runoff and infiltration processes; The goal of the first study was to simulate the long-term (18,000 yrs) multi-phase (liquid and vapor) water fluxes and associated chloride fluxes in the northern Mojave Desert by applying different reconstructed boundary conditions in the simulation. The results showed that the observed near-surface chloride peak reflected the combined boundary conditions of precipitation, root-water uptake, and soil evaporation. The results showed that climate shift alone (with normal precipitation patterns) was not the major driving force that initiated the observed near-surface chloride accumulation. Rather, the results showed that root water uptake and extreme storm events, embedded within the normal precipitation patterns, were the major driving forces that controlled the paleo-water fluxes and chloride profile distributions. Also, the results showed that chloride accumulations were highest at the zone of maximum root zone distribution of Mojave Desert shrubs, not the depth of the roots. Thus, observed chloride accumulations deeper than the active root zones still cannot been fully explained; The second study was to assess three different methods used to generate spatially distributed hydraulic properties, by simulating surface runoff on a semi-arid rangeland at the Walnut Gulch Experimental Watershed, outside of Tombstone, AZ. By collecting 66 soil samples (2 samples at each of 33 sites) and using pedotransfer functions, soil hydraulic properties were derived. Then three methods were used to generate the parameter fields of a two-dimensional diffusion wave model to simulate a total of eight storm events with measured runoff. The results showed that co-kriging was the best approach to represent the spatial variability of soil hydraulic properties. The results also showed that the need to calibrate plant interception models based on historical records of shrub versus grassland coverage; The goal of the third study was to understand the influence of desert pavement on infiltration and surface runoff, and to calibrate relevant Green-Ampt infiltration parameters. To achieve the goal, twelve rainfall simulator tests were conducted in the Mojave National Preserve, CA and the in-situ infiltration and surface runoff were measured. The results showed no statistical difference between the infiltration characteristics between plots with and without desert pavement (i.e., clast) surfaces. The results indicated that variability of soil texture exerted a larger effect on infiltration than the effects introduced by the surface clasts only. However, an optimization method was necessary to calibrate the Green-Ampt parameters. In these cases, the optimized parameters underestimated and overestimated hydraulic conductivity values, compared to pedotransfer functions and tension infiltrometer tests, respectively; The modeling results in this dissertation showed how numerical simulations can be used to assess soil moisture dynamics and model parameter variations in arid and semi-arid regions. The studies quantitatively modeled several hydrologic processes in the northern Mojave Desert such as vertical water fluxes, and helped determine effective hydraulic conductivities on alluvial fans with desert pavement. These results provide fundamental knowledge of infiltration and deep percolation in this desert region of the United States, and show how features of these sites were well preserved during the physically-based modeling processes. We showed also that the modeling approaches can more effective than empirical correlations for predicting water flux as environmental (i.e., climate) conditions change. But it is noted that appropriate boundary conditions and model parameters were found to be most important aspects of producing reliable modeling results.


Hydraulic; Impact; Infiltration; Moisture; Proce; Processes; Properties; Rainfall; Runoff; Rainfall-runoff; Soil; Soil Moisture; Spatiotemporal; Variation

Controlled Subject


File Format


File Size

2222.08 KB

Degree Grantor

University of Nevada, Las Vegas




If you are the rightful copyright holder of this dissertation or thesis and wish to have the full text removed from Digital Scholarship@UNLV, please submit a request to and include clear identification of the work, preferably with URL.