|Science Results (Winter 2014)|
Langhans, C. University of Melbourne
Infiltration and runoff rates on rangelands can vary considerably due to a number of interrelated factors including soil and plant characteristics, slope gradient and length, microtopography, and rainfall rates and duration. This study addresses the effects of runoff flow depth and microtopography on infiltration rates in a shrub dominated ecosystem. The concept is that the infiltration rates under the shrubs are higher than in the areas between the shrubs. Because the shrubs occur on mounds, this means that as the runoff flow depth increases, the area covered by water includes topographically higher areas that have higher infiltration rates thus increasing the area averaged infiltration. A model was developed that combines the above concept with partial contributing area under variable rainfall intensity and the widely used Green-Ampt infiltration equation. The performance of the new model was compared with that of a conventional Green-Ampt equation using plot and small watershed rainfall runoff data from the USDA Walnut Gulch Experimental Watershed in south eastern Arizona. The new model reproduced hydrographs better for the plot data while both models did equally as well for the small watershed data. The proposed model offers some practical and theoretical advancement towards a less scale dependent way of modelling runoff at the hillslope scale.
Moran, M.S. Southwest Watershed Research Center
Grasslands across the United States play a key role in regional livelihood and national food security. Yet, it is still unclear how this important resource will respond to the regional drying and warming predicted with climate change. In this study, we investigated the impact of the early 21st century drought on six grasslands across the southwestern United States. We reported an exceptional decrease in grassland growth, and the replacement of native grasses with less-nutritious and more-fire-prone invasive grasses. By providing a model of grasslands’ response to prolonged warm drought and an operational means to monitor it, this analysis has important implications. First, ranchers have important new information for managing grasslands under predicted climate change to lower fire risk, minimize loss of forage, and retain ecosystem services. Second, the aberrant behavior reported here suggests that grasslands can serve as an early indicator of impending climate change. These compelling results in a natural setting at the regional scale should play a role in future grassland research, management and policy.
Broxton, P. University of Arizona
Fast responding headwater river basins and small streams pose a significant threat to life and property throughout the United States. The National Weather Service (NWS) of the National Oceanographic and Atmospheric Administration (NOAA) is responsible for providing flash flood warnings to the public. In this study, the NWS, in cooperation with the USDA-ARS, and the University of Arizona, developed a modeling system (KINEROS/hsB-SM) to treat runoff generated from fast responding overland flow, slower responding groundwater storm response, and runoff generated from snowmelt and rain on snow. It also includes the ability to incorporate high-resolution weather and radar data from the National Weather Service. The model is currently being used in an operational evaluation at the National Weather Service Binghamton, New York Weather Forecast Office. The evaluation demonstrated that KINEROS/hsB-SM can be successfully used for flash flood forecasting in the area. The KINEROS/hsB-SM modeling system provided valuable information on the size and timing of flash flood which is not available using current operational NOAA/NWS flash flood forecasting methodologies.
Sanches Oliveira, P. University of São Paulo
The tropical forests, savannas, biological diversity and freshwater resources of Brazil are among the world’s largest. The Brazilian Cerrado is considered one of the most important Brazilian biomes, covering an area of 2 million km2 (~22% of the total area of Brazil) and is the second largest biome in South America. The physiognomies of the Cerrado vary from grassland to savanna to forest. Because of its endemic plant and vertebrate species, this biome has been classified as one of the 25 global biodiversity hotspots. Despite the importance of the Brazilian Cerrado, knowledge of the water balance dynamics associated with the change from native cover to grassland and cropland is still limited. In this study, we review the available literature on the components of the water balance in the Brazilian Cerrado to obtain a better understanding of the consequences of deforestation for the water balance in the Cerrado. Results indicated that deforestation and streamflow, erosion processes and sediment yield have tended to increase. We detected no statistically significant decrease in the overall annual average evapotranspiration in the Cerrado biome in the 2000 to 2010 period, but we did detect significant decreases in some states with rapid deforestation.
Polyakov, V.O. Southwest Watershed Research Center
This study looks at the potential for using low-tech, rock check dams as runoff and erosion control structures in a semi-arid environment. We used two small watersheds that have been monitored for surface water storm runoff and sediment export over the last 3 decades. The rates and amounts of surface water runoff were not appreciably changed after the installation of the check dams, however the structures did retain a significant amount of sediment over the four year period of the experiment. After the experiment the dams were approximately 80% filled to capacity for sediment retention. We will continue to monitor the experiment in the future, particularly looking for the potential for the depositional areas above the structures to act as a reservoir in retain more water than would otherwise be retained, thus allowing for the establishment of vegetation in the channels. Check dams have the potential to make a difference in sediment movement in the semiarid environment, at least over the short term, if properly installed and maintained.
Gao, P. University of Syracuse
This study looks at the sediment that is transported from watersheds ranging in size from approximately one acre up to 50 square miles and located in southern Arizona. We looked at both the rate that sediment was exported from the watersheds during a rainfall event, as well as the total amount of sediment exported during entire events. The changes in sediment export rates during events did not follow any common pattern, which indicates that a multiplicity of processes occur at different times and with different storms, resulting in complex sediment responses. However, we did find that the total amount of sediment exported during entire events was more or less linearly related to the total amount of surface water runoff volumes. Finding this simple relationship between runoff volumes and sediment exports might be an important tool for general characterization of differences across geographic regions as well as in easily identifying general trends over time, such as might potential changes associated with climate change.
Socha, L.M. State University of New York (SUNY)
Nitrogen (N) and phosphorus (P) are two important limitations to grassland productivity. However, how N and P affect the physical and biological features regulating atmosphere/biosphere carbon dioxide exchange, an important component of plant growth and productivity, are not clear. We measured seasonal leaf area index (LAI) net annual productivity (ANPP in g per year) and net ecosystem carbon dioxide exchange (NEE), and its constituent fluxes of ecosystem respiration (Re) and gross ecosystem carbon dioxide exchange (photosynthesis, or GEE) in a full factorial field study with N+, P+, N+P and unfertilized control plots of temperate grassland to elucidate if these nutrients co-limit productivity and ecosystem carbon balance. We found that only N-fertilization significantly increased ANPP, LAI and all carbon fluxes, while addition of P did not change these variables in any way. N-fertilization stimulated rapid growth and LAI, inducing stronger carbon-sink activity in these plots, but also resulted in strong depletion of soil moisture, especially later in the season when rains were more limited. This water limitation and self-shading from extensive canopy growth likely limited LAI-corrected GEE (GEEa), which was markedly lower in N-fertilized plots. We concluded that N- and P-influences on ecosystem processes are seasonally dynamic, and by differentially influencing coupled aboveground and belowground biological activity, soil N- and P-availability may interact to reduce the strength of grasslands to act as carbon sinks.
Nelson, E. Bowdoin College
Ecosystems, and the biodiversity and services they support such as providing food, clean water, and clean air, among many, are largely dependent on climate. During the 20th century, climate change has had documented impacts on ecological systems, and these impacts are expected to increase as climate change continues and perhaps even accelerates. This paper synthesizes a body of scientific studies of the way climate change is affecting biodiversity, ecosystems, ecosystem services, and what strategies might be employed to decrease current and future risks. Building on past assessments of how climate change and other stressors are affecting ecosystems in the United States and around the world, we approach the subject from several different perspectives. People experience climate change impacts on biodiversity and ecosystems as changes in ecosystem services; people depend on ecosystems for resources that are harvested, their role in regulating the movement of materials and disturbances, and their recreational, cultural, and aesthetic value. Thus, we review newly emerging research to determine how human activities and a changing climate are likely to alter the delivery of these ecosystem services.
Morris, C.E. USDA-ARS
Invasions by exotic plant species can generally be described with a logistic growth curve divided into three phases: introduction, expansion and saturation. This model is constructed primarily from regional studies of plant invasions based on historical records and herbarium samples. The goal of this study was to compare invasion curves at the site scale to the predicted logistic growth curve using long-term datasets that documented exotic weed invasions.
Five datasets ranging from 41-86 years in length were recovered from five sites in four western states. Data for the following seven exotic species were analyzed using regression analysis to evaluate goodness of fit to a logistic curve: crested wheatgrass, cheatgrass, lehman lovegrass, desert madwort, saltlover, russian thistle and tall tumble mustard. Of these species, six increased in abundance over time in at least one location and five had R2 > 0.90 when fitted to a logistic curve. All other species/location combinations were characterized by sporadic spikes and crashes. One species that increased over time (lehman lovegrass) did not exhibit a lag phase at all, suggesting that in some cases, seedbank analysis may be a better predictor for invasion risk by exotic plants. Exotic forage species selected for their suitability to their respective regions exhibited medium and long introduction phases, suggesting that colonization processes are at least as important as evolutionary processes during invasion. We conclude that long-term datasets are an under-utilized tool that may provide invaluable information for studying the invasion process of exotic plants.
Barro-Gafford, G.A. University of Arizona
The conversion of many historic grasslands to shrublands and savannas worldwide has the potential to alter how ecosystems will respond to global climate change because the dominant plants, either grasses and woody plants, have different responses to these changes. We explored how temperature and precipitation control carbon fluxes within a pair of semiarid shrublands in Arizona that had undergone woody plant expansion. We found significant differences in the functioning of the two plant functional types, in that the shrubs were able to consistently conduct photosynthesis across a broader temperature range than co-occurring grasses during dry periods, but with grasses outperforming shrubs during the wetter monsoon season. Landscape position modulated these temperature sensitivities, as the range of functional temperatures and maximum rates of photosynthesis were two to three times greater within the riparian shrubland in dry times. Given projections of more variable precipitation and increased temperatures, differences in physiological activity within these growth-forms are likely to drive patterns of ecosystem carbon exchange. As access to more stable groundwater declines with decreased precipitation input, these differential patterns of temperature sensitivity among growth-forms dependent on connectivity to groundwater will only become more important in determining whether these ecosystem will absorb more or less atmospheric carbon dioxide in the future.
Nagler, P.L. USGS
River basins frequently support both irrigated agriculture and natural vegetation along rivers. To accurately account for the water used by the crops and natural vegetation, new methods incorporating readily-available satellite data are needed. We developed such a method using freely and globally available satellite data from NASA in combination with local weather data. The algorithm capably predicted crop and natural riparian vegetation water use (termed evapotranspiration) and then further validated over five different irrigation districts worldwide with predictions within 10% of measured results in each case. The algorithm provides more accurate information to farmers and land managers for improved water resource management.
Sugg, Z. University of Arizona
Population in-migration related to environmental amenities in Arizona and many other western regions have been associated with high rates of urbanization. Impervious areas created by unbanization can produce significant changes in hydrological processes by altering runoff, increasing peak flows and degrading water quality. Mapping impervious surfaces with remote sensing techniques is an effective way to quantify impervious cover and thereby improve understanding of the impacts of urbanization on runoff processes. The most common approach using remotely derived vegetation indices, however, is problematic in arid and semiarid environments where vegetation is patchy and often senescent. This paper describes a method for mapping impervious surfaces using high-resolution imagery for an urbanizing semi-arid area. At the neighborhood scale we found that the automated method of image classification creates impervious area maps that are comparable in accuracy to manual delineation methods. The same method was then adapted to produce a map of impervious surfaces in the entire city of Sierra Vista, Arizona and the surrounding subwatershed with high accuracy. This method overcame a common limitation of high spatial resolution imagery, which is that broad spatial coverage can be difficult due to the high amount of data per image.
Yang, Y. Tsinghua University, China
Evapotranspiration (ET) is a major component of the global water cycle, but it is difficult to measure. Furthermore, even when such sophisticated measurements are available, information about how much ET is from bare soil evaporation (E) and how much is from plant transpiration (T) is often desired to better understand the water use of ecosystems. In this paper, an algorithm that combines routine satellite measurements with ground-based ET measurements to partition ET into its respective components was tested with measurements of E and T made in a desert shrubland environment in Arizona where it was relatively easier to measure E and T separately. The algorithm capably produced accurate estimates of E and T at the time of the satellite measurements, though challenges about how to interpolate the estimates between satellite measurements which are available once every few days to weeks remain.
Ely, C.C. Environmental Protection Agency (EPA)
Long-term land-use and land cover change and their associated impacts pose critical challenges to sustaining vital human and environmental benefits provided by healthy watersheds. In this study, a methodology was developed to use real estate parcel data to evaluate land use trends in southeast Arizona’s San Pedro River Watershed. Changes to parcel size are examined decade by decade, for two intervals: from 1882 to 2012, and from 1971 to 2012. Graphs and maps were then developed to illustrate the changes in both parcel size and parcel density for both intervals. The parcel density maps and graphs employ housing density categories developed by the Environmental Protection Agency’s Integrated Climate and Land use Scenarios project. The purpose of this study is to 1) improve and describe a methodology for evaluating land use trends using parcel data; 2) display land use trends in a portion of the San Pedro Watershed using parcel data; and 3) discuss the implications of the analysis for evaluating environmental impacts with modeling tools. In addition it aids in assessing indirect effects of development as required by the National Environmental Policy Act.
Singh, R.K. ARTS (Contractor for USGS)
Accurately estimating the amount of water lost to evaporation in the Colorado River Basin (CRB) is important for assessing and managing limited water resources in the basin. We have developed a detailed, field-level, map of evaporation for the CRB for water use assessment. To do this, a computer model was developed that incorporated meteorological and satellite measurements for the year 2010. Our results show that cropland had the highest evaporation among all land cover classes except for open lake or river water. Comparisons of the model estimates with actual evaporation measurements and basin water budget measurements indicated a good agreement. Likewise, our model estimates agreed well with coarser-scale model estimates, but our map is more suitable for water use assessments at the agricultural field level.
Hamerlynck, E.P. Southwest Watershed Research Center
All desert systems experience predictable, long-lasting dry spells, yet, despite their prevalence, little attention has been given to ecosystem carbon dioxide fluxes in deserts over annual drought periods. We continuously measured soil surface respiration under and between shrubs, soil carbon dioxide concentrations and fluxes across a range of soil depths, and ecosystem carbon dioxide flux in a Chihuahuan Desert shrubland over a three month dry period before the 2012 summer monsoon season. We found that under these dry conditions, soil respiration rates were often negative over the night, indicating soil carbon uptake. Nocturnal soil carbon uptake depended on how strong soil to air temperature differences were, and these conditions were more common in exposed intercanopy locations than under shrubs. While soils were taking up carbon dioxide, our soil CO2 profiles showed that CO2 at 2cm was actually below atmospheric levels, which was likely due to soil carbonate dissolution. In the morning, soil respiration rates became positive, but did so to a greater degree and faster than fluxes at different soil depths. This “out-gassing” is probably due to CO2 at and just below the soil surface being released as carbonate precipitates out. This process is probably aided by atmospheric turbulence that occurs as the air and soil surface heat up in the morning, and resulted in much higher soil respiration rates than would be predicted from prevailing temperatures. These results show inorganic carbon dynamics are the dominant feature in ecosystem gas exchange over annually re-occurring drought, and these will likely respond strongly to continued atmospheric carbon dioxide increases, as well as the warmer and drier pre-monsoon periods predicted across the Southwest U.S.
Stillman, S. University of Arizona
Arid and semi-arid regions account for approximately one-third of the land mass of earth. These regions are experiencing continued pressure from population growth in many parts of the world. Water is a critical resource in these regions and is often in short supply. Detailed study of precipitation and storm patterns is important if we are to continue to populate and use these regions. This paper analyzed the highly detailed long-term rainfall records from the USDA-ARS Walnut Gulch Experimental Watershed (WGEW) near Tombstone, Arizona during July and August when the majority of rainfall occurs from monsoon thunderstorms. The study found that there were no significant trends in rainfall total, intensity or frequency of storms from 1956 to 2011. However, the multi-decadal variation in storm intensity and spatial coverage of storms was correlated with the Pacific Decadal Oscillation (PDO) and the Atlantic Multi-decadal Oscillation (AMO) indices. The rainfall patterns were found to have a pronounced diurnal cycle with the highest storm totals and frequency occurring between 3:00 and 10:00 p.m. It was also found that 5-8 rain gauges (of the 88 present in the watershed) can provide a reasonable estimate of the area-averaged monthly total precipitation during July and August. This has important implications for resources needed to adequately monitor and model rainfall in this environment.
Brand, A. University of Arizona
Freshwater ecosystems, and the habitat they support are relatively rare in the semi-arid southwest and significant effort and expense is being directed at restoration of these systems. In the Middle Rio Grande, the lack of flooding due to river impoundments has contributed to the spread of exotic vegetation with dense understory fuel loads. Restoration has focused on understory vegetation thinning but it is unclear how these actions impact bird populations. We quantified densities of five guilds of birds (canopy, midstory, and understory nesting birds; water-dependents; and spring migrants) across 12 vegetation composition-structure types. We then estimated bird population changes expected from restoration scenarios for three possible management options (mechanical clearing, hand thinning, and wetland restoration). The estimated projections of bird population changes will help managers evaluate biological impacts of restoration measures being applied on the Middle Rio Grande. This study is part of a larger effort to link science with valuation of non-market ecosystem services in semi-arid riparian systems.
Cendrero-Mateo, M.P. University of Arizona
Water deficit limits plant photosynthesis and decreases crop yield. An important challenge is to establish a rigorous landscape-scale indicator of plant photosynthesis. The emission of fluorescence from plant chlorophylls provides direct measure of plant photosynthesis. Chlorophyll fluorescence can be measured at both field and landscape-scale using satellite-based sensors. However, the relationship between plant photosynthesis and chlorophyll fluorescence changes with water availability and light intensity. In this study, we determined that at ambient growth light conditions chlorophyll fluorescence is directly correlated with photosynthesis and it can be use as a water-deficit indicator. The results from this study provides quantitative support for the proposed observation time (09:30-10:00 AM) of the upcoming European Fluorescence Explorer (FLEX) satellite mission for monitoring plant responses to water deficit, as well as a promising starting point for interpretation of seasonal changes in crops growth using satellite-based sensors.
Gaunter, L. Institute for Space Sciences, Free University of Berlin, Germany
With increasing food demand, it is critical to know the status of global food supply. A new satellite has just been launched to make worldwide measurements of cropland productivity. Based on these measurements, we found that the US Western Corn Belt is producing at a rate 50–75% higher than previously predicted, and in fact, is the highest producing ecosystem on the planet. Our results contribute to the understanding of the global carbon cycle in general, and to food security in particular.
Nichols, M.H Southwest Watershed Research Center
No measurements can be perfectly accurate. Runoff is measured on the Walnut Gulch Experimental Watershed in southeastern Arizona at ponds located at the outlet of subwatersheds. Uncertainty analyses were conducted to review current measurement procedures and to understand the effects of instrumentation, field methods, and data processing procedures on the accuracy of measured runoff. The largest measurement uncertainty was associated with flows that were less than 50 m3 in magnitude and with flows that exceeded the pond capacity and resulted in outflow through spillways. The largest source of measurement uncertainty was found to be related to establishing the position of the float sensor relative to the pond outlet elevation. This analysis is of practical importance for improving field measurement methods and for assessing the quality of runoff predictions that are based on simulation models developed using measured runoff data.
Oliveria, P. Universidad De Sao Paulo
The most widely used method to estimate runoff amount from storm rainfall is the Natural Resources Conservation Service (NRCS) Curve Number (CN). The CN is obtained either from NRCS tabulated values based on soil and vegetation type, landuse, and management of an area of interest or from calculations from rainfall-runoff data. The two methods were used in an experiment in south central Brazil that measured runoff from natural rainfall on large plots with native forest vegetation (termed Cerrado in Brazil), bare soil, pasture, soybeans, millet, and sugarcane. The CN for each of the plot cover types was obtained using the standard tabular values developed by NRCS and five different calculation methods. The results showed that there was no difference between observed runoff and predicted runoff using the tabular and calculation methods for the cropland plots but that none of the methods could predict reasonable runoff amounts for the forest plots. These results for the forest plots are similar to those found in the US and are attributed interception of rainfall, mainly by litter on the forest floor and secondarily by the forest canopy. Because the Cerrado is an important biome in a large part of Brazil, it will be necessary to develop new experimental procedures to account for the interception losses in the CN method.
Barron-Gafford, G.A. University of Arizona
Soil respiration is a significant contributor to atmospheric carbon dioxide concentrations. In order to accurately predict how much carbon dioxide will be respired in the future, critical information about how soil respiration responds to environmental conditions is needed. We carried out measurements of soil respiration and potential environmental drivers in a grassland habitat in southern Arizona and found that rates of soil respiration responded both to moisture in the soil and photosynthesis rates in nearby plants. But, this response differed depending on whether the nearby plant was a bunchgrass or a tree. Thus, improvement in our predictions of the carbon cycle may warrant the inclusion of plant-type specific responses that we found.
Guillod, B.P. ETH, Zurich, Switzerland
Precipitation falls from the atmosphere and replenishes moisture in the soil. This moisture in turn can evaporate and return to the atmosphere, possibly leading to more precipitation. This link has long been a topic of research because of its potential to improve weather forecasting. This study investigated the poorly understood link between surface evaporation and subsequent precipitation by investigating the impact of before-noon evaporation strength on the frequency of afternoon precipitation over the contiguous US. We analyzed satellite and radar data products, met station data, and computer model output. While most computer model datasets agree on there is a positive relationship between evaporation strength and precipitation in the eastern and southwestern US, observation-based estimates also indicated a strong relationship in the Central US which was not found in computer model analysis. Further analysis reveals that precipitation events tended to be clustered in space and time and this leads to a relationship which may not be casual. This study’s findings uncover a large range of unresolved issues for this topic. More investigations are needed to better pin down this complicated relationship
Stillman, S. University of Arizona
Knowing the amount of moisture in the soil is important for many different applications including meteorology, agriculture, and water management. Routine measurements of soil moisture are uncommon and, even when available, are generally not available over a wide area of interest such as a field or watershed. The purpose of this study was to take advantage of the unique long-term, spatially-distributed, soil moisture and precipitation dataset that has been collected over the past decade over the USDA-ARS Walnut Gulch Experimental Watershed to develop a method to determine soil moisture using more commonly available precipitation data alone. This model is then used to effectively increase the amount of soil moisture observations from the 13 to 17 normally available in any given year to the 37 to 88 available precipitation measurement sites and extend the time period of measurements from 10 (2002-2011) to 56 (1956-2011) years. The daily soil moisture over the watershed has a large spatial range following a summer thunderstorm and a small range during dry drought-like periods. Spatially and seasonally averaged soil moisture can be estimated nearly as well with the 13-17 soil moisture measurement sites as with samples at all rain gauge locations; however, estimates of daily spatial variability can be significantly hindered using fewer samples. The new longer-term and higher-resolution soil moisture dataset can be useful for new experimental applications such as hydrological modeling, soil moisture-sensing satellite calibrations, and plant growth models.
Scott, R.L. Southwest Watershed Research Center
The vegetation composition of many ecosystems around the world and in the U.S. is rapidly changing. The causes of these rapid changes are intimately connected with climate change and management practices, but the consequences of these shifts in composition and structure on the fundamental ecosystem services of water and carbon cycling are not well understood. For this study, we use multi-year hydrological and meteorological data to evaluate how soil water accessibility affects the magnitude and variability of the biosphere-atmosphere exchanges of water and carbon dioxide amongst three proximate ecosystems that are representative of varying degrees of mesquite tree invasion. We found that groundwater access increased in ecosystems with greater amounts of trees rather than grasses, and that groundwater accessibility led to significant changes in magnitude and variability in ecosystem water and carbon cycling. Even as depth to groundwater increased from grassland to shrubland to woodland, the woodland had the highest evapotranspiration in excess of precipitation and the grassland was lowest. How the greater density of trees affected the amount of carbon taken in by an ecosystem was more complex. Woodland carbon sequestration was the largest but surprisingly similar to the less mature and dense shrubland, and both systems took in much more carbon than the grassland. Vegetation change in areas where the accessibility to deeper soil water increases will likely increase carbon sequestration but at the expense of higher water use.
Zhang, Xin Beijing Normal University
Many regions are suffering from severe and frequent droughts, but their impacts on the grasslands and forests around the world are still uncertain. In this study, we investigated the impacts of the prolonged drought during the early 21st century on four ecosystems in China including forests, savannas and grasslands. Our results indicate that the water use of plants in all these ecosystems is affected by not only the current drought conditions, but also the drought conditions of the previous year. Based on these results, we developed and validated a simulation model to predict the effect of drought on vegetation growth. This model is particularly useful for understanding the impact of prolonged drought on our extensive forests and grasslands. It will help managers plan for the impact of predicted climate change on valuable natural resources, especially if prolonged drought continues in the future.
Nguyen, U. University of Arizona
The Upper San Pedro River is one of the few remaining undammed rivers that maintains a vibrant riparian ecosystem in the southwest U.S., but its riparian forest is threatened by diminishing groundwater and surface water inputs. To assess the temporal and spatial trends in San Pedro riparian health, we used satellite measures of vegetation abundance and vigor from 1984 to 2012. The river was divided into a northern reach, which has mainly intermittent and ephemeral flow reaches, and a mainly perennial flow, southern reach. The satellite measures of riparian greenness showed a 20% drop for the northern reach and no net change for the southern reach. Vegetation abundance and vigor were positively related to river flows, which decreased over the study period in the northern reach, and negatively related to increasing trends in reach air temperatures, which have increased by about 1.4 degrees Celsius from 1904 to 2012. Hypothesized increases in riparian and upland evaporation could not explain the reduction in river flows because the satellite measures of vegetation health decreased in both the riparian corridor and in the uplands over the 1984-2012 period. Regional groundwater pumping is a feasible alternative explanation for decreased flows and deterioration of the riparian forest in the northern reach.