Skip to main content
eScholarship
Open Access Publications from the University of California

Scholarly Works (33 results)

  • Thesis
  • Peer Reviewed
UC Davis Electronic Theses and Dissertations (2021)

Due to population growth and progress made in sustainable water management, water resources availability to California’s croplands is becoming more scarce. Besides, due to climate change, year-to-year water supplies are becoming increasingly unpredictable. As 80% of all human use of water is diverted for farmland irrigation, spatial estimates of crop evapotranspiration with high accuracy, from the field to watershed scale, have become increasingly important for water resources management. The present work aims to improve and evaluate an algorithm for estimating spatial crop evapotranspiration and explore its application to water resource management. The generated knowledge and tools from this study can provide data-driven guidance for water resources planning at the watershed level, contributing to a more sustainable use of water in California’s agricultural regions. The first chapter focused on improving and evaluating a semi-empirical model to map crop evapotranspiration at a 30m resolution using Landsat satellite images and concurrent spatial meteorological data. The algorithm was optimized and cross-validated with available field measurements from over 22 sites in California’s Central Valley for major crops, including alfalfa, almond, citrus, corn, pasture, and rice. The evapotranspiration estimates explained more than 70% variance in daily measurements with an RMSE of 0.88 mm day-1. The site-level evaluation was repeated with the estimates from another remote sensing algorithm, PT-JPL, which is developed by scientists at the NASA Jet Propulsion Laboratory. The two algorithms share a similar theoretical basis and design, and an improvement over a PT-JPL was observed when the proposed approach is crop-specifically calibrated. This chapter demonstrated that land use information and field measurements are useful for reducing the uncertainty of spatial crop evapotranspiration estimates. The model presented in the first chapter can be applied for continuous monitoring of the dynamics and spatial heterogeneity of crop consumptive water use in California at a field scale. Chapter 2 scaled up the proposed algorithm to estimate annual evapotranspiration at 30m resolution across the entire Central Valley for water years 2014 and 2016. It examined the spatial pattern of crop water consumptive use at a regional level and a local scale of Groundwater Sustainability Agencies (GSAs). We estimated a mean (x̅) evapotranspiration of 820 mm yr-1 with a standard deviation (σ) of 290 mm yr-1in 2014. Compared to 2014, the total estimated evapotranspiration in 2016 increased by 9.6%, mostly because of land-use conversion from fallow/idle land to cropland. Large variation of agricultural water use was found across different and within the same crop types, with a coefficient of variation ranging from 8% for Rice (x̅ = 1110, σ = 85 mm yr-1) to 59% for Pistachio (x̅ = 592, σ = 352 mm yr-1). Besides, tree crops, such as Almonds, were found to have a large variation of evapotranspiration among and within GSAs. To further understand the spatial variability of almond’s evapotranspiration at the GSAs scale, an empirical model using normalized difference vegetation index (NDVI) and orchard age was optimized and evaluated. Orchard age was shown to be an important factor contributing to the variability of evapotranspiration across GSAs. However, spatial variation of other physical factors that could affect NDVI, such as salinity at the root zone, will likely have stronger control over almond’s evapotranspiration, explaining its spatial variation within and among GSAs. Overall, this chapter demonstrates how spatial estimates of evapotranspiration can be used to provide data-driven guidance for more effective land use and water planning across regional and local GSA scales. In chapter 3, the proposed algorithm was implemented to estimate daily crop evapotranspiration at 30m resolution in the Pajaro Valley, CA for 2009 and 2011. The objective of the study was to develop and evaluate methods to estimate agricultural applied water from field to local agency scales. To estimate applied water at the field scale, the daily spatial evapotranspiration was coupled with soil water balance for simulation of daily fluxes in a water balance at 30m resolution. This coupled model requires a crop-specific optimization on application efficiency, using a limited number of field-scale applied water records. When the model estimates are spatially aggregated by fields and temporally aggregated to the annual total, the evaluation showed that our applied water estimates agreed well with the pumping records at the field scale, explaining 86% of variability with an RMSE of 39.6 million liters. When aggregated over the Pajaro Valley, we estimated a total of agricultural groundwater pumping ranging from 53.9 in 2009 to 42.6 billion liters in 2011, which deviated from the sum of all agricultural pumping and surface water delivery by less than 4.6%. Over 16 subregions in the valley, our estimates explained more than 89% variance in pumping records with an RMSE of 1.2 billion liters. Our estimates of water balance fluxes were also compared with that by USGS, to evaluate the model’s ability in representing the hydrological process of the Valley. Results from chapter 3 demonstrated that remote sensing evapotranspiration estimates can be used to estimate regional agricultural applied water with reasonable accuracy yet in a cost-efficient way, over areas where only a limited number of field-scale applied water data is available for calibrating a soil water balance model.

    Cover page: Remote Sensing of Evapotranspiration in California Cropland: Improvement and Hydrological Applications
    • Thesis
    • Peer Reviewed
    UC Davis Electronic Theses and Dissertations (2022)

    The increasing trend of active wildfire events in California poses risks to both wildland ecosystems and human communities, causing substantial ecological damages and economic losses. However, the mechanism of wildfire damage within vegetated landscapes and human settlements has been understudied, especially for the effects of changing climates and the expanding wildland urban interfaces. Recent advances in remote sensing and machine learning enabled fine-scale observation of land changes from wildfire attacks and an improved analysis of environmental and anthropogenic factors controlling the wildfire damage. This dissertation has sought to bridge the knowledge gaps of wildfire risks across the land through remote sensing and machine learning approaches. Specifically, this research has aimed to 1) disentangle effects of environmental factors on burn severity under various extreme climatic conditions in California’s northern coastal mountains and predict near-real-time burn severity from remote sensing images and geospatial data; 2) develop an accurate and effective framework of biannual analysis of building footprints and wildland urban interface mapping from high resolution aerial photos; 3) perform the first comprehensive analysis of factors determining building damage from wildfire events at the daily scale across California. The results of this dissertation highlighted the importance of targeting areas with high burn severity or building damage risk for fire adaptation and mitigation strategies in a changing climate and expanding human settlement. These findings, when combined with building footprints and wildland urban interface products developed through a modeling framework, contribute to the guidance of land use planning to create fire- adaptive and resilient communities in California.

      Cover page: Remote Sensing of Land under Wildfire Risks: Evaluation of Burn Severity, AI-Driven Mapping of Wildland-Urban Interface Structures, and Analysis of Building Damage from Wildfires in California
      • Article
      • Peer Reviewed
      UC Irvine Previously Published Works (2014)
      Aim: Precipitation controls the production of semi-arid plants through various mechanisms that operate at a range of time-scales. Short-term variation in precipitation affects vegetation through adjustments in plant physiology and leaf phenology, whereas long-term effects are mediated by plant establishment and mortality, community composition and disturbance regimes. Our goal is to use remote sensing observations to separate the short- and long-term effects of variation in precipitation on ecosystem production. Location: California, USA. Methods: We used time series of gridded absorbed photosynthetically active radiation (APAR) to quantify the short- and long-term responses of diverse ecosystems to variation in precipitation across large productivity and precipitation gradients. We investigated the relationships between temporal sensitivity of APAR to interannual variation in precipitation and mean annual precipitation (MAP), ecosystem properties and disturbance. Results: APAR increased with precipitation both interannually within locations and across locations with MAP. The slope of the interannual relationship, which reflects the sensitivity of APAR to short-term fluctuations in precipitation, varied with climate, vegetation type and structure, and time since disturbance. The interannual APAR sensitivity decreased from c. 0.5 MJ m-2mm-1 at a MAP of 300mm year-1 to less than 0.05 MJ m-2mm-1 at 1000mm year-1. The slope of the spatial relationship, which reflects the long-term sensitivity of APAR to climate, decreased from c. 2.5 MJ m-2mm-1 at 300mm year-1 MAP to c. 0.6 MJ m-2mm-1 at 1000mm year-1. The initial physiological and leaf area effects of a precipitation shift were amplified five-fold over time by gradual changes in population density and species composition. Main conclusion: The impact of a hydroclimatic shift on the primary production, structure and function of California's terrestrial ecosystems depends heavily on time-scale and how rapidly changes in plant population density and community composition can occur. © 2013 John Wiley & Sons Ltd.
      • 1 supplemental PDF
      Cover page: Ecological consequences of variation in precipitation: separating short‐ versus long‐term effects using satellite dataCreative Commons 'BY' version 4.0 license
      • Article
      • Peer Reviewed
      Faculty Publications (2008)
      We assessed the multidecadal effects of boreal forest fire on surface albedo using Moderate Resolution Imaging Spectroradiometer (MODIS) satellite observations within the perimeters of burn scars in interior Alaska. Fire caused albedo to increase during periods with and without snow cover. Albedo during early spring had a mean of 0.50 ± 0.03 for the first three decades after fire, substantially higher than that observed in evergreen conifer forests (0.34 ± 0.04). In older stands between 30 and 55 years, albedo showed a decreasing trend during early spring, probably from a growing spruce understory that masked surface snow and caused increases in both simple ratio (SR) and enhanced vegetation index (EVI). During summer, albedo decreased by 0.012 ± 0.005 in the year immediately after fire (from 0.112 ± 0.005 to 0.100 ± 0.010). In subsequent years, summer albedo increased rapidly at first and then more gradually, reaching a broad maximum in 20–35 year stands (0.135 ± 0.006). These measurements provide evidence for a well-developed deciduous shrub and tree phase during intermediate stages of succession. Averaged over the first 5 decades, shortwave surface forcing from fires was −6.2 W m−2 relative to an evergreen conifer control and −3.0 W m−2 relative to a control constructed from 2000 to 2003 preburn observations. These forcing estimates had a magnitude substantially smaller than previous estimates and suggest that, at a regional scale, evergreen conifer stand density may be lower than that inferred from chronosequence studies.
      • 3 supplemental files
      Cover page: Changes in surface albedo after fire in boreal forest ecosystems of interior Alaska assessed using MODIS satellite observationsCreative Commons 'BY' version 4.0 license
      • Article
      • Peer Reviewed
      UC Irvine Previously Published Works (2022)
      Major wildfires and heatwaves have begun to increase in frequency throughout much of the United States, particularly in western states such as California, causing increased risk to public health. Air pollution is exacerbated by both wildfires and warmer temperatures, thus adding to such risk. With climate change and the continued increase in global average temperatures, the frequency of major wildfires, heat days, and unhealthy air pollution episodes is projected to increase, resulting in the potential for compounding risks. Risks will likely vary by region and may disproportionately impact low-income communities and communities of color. In this study, we processed daily particulate matter (PM) data from over 18,000 low-cost PurpleAir sensors, along with gridMET daily maximum temperature data and government-compiled wildfire perimeter data from 2018-2020 in order to examine the occurrence of compound risk (CR) days (characterized by high temperature and high PM2.5) at the census tract level in California, and to understand how such days have been impacted by the occurrence of wildfires. Using American Community Survey data, we also examined the extent to which CR days were correlated with household income, race/ethnicity, education, and other socioeconomic factors at the census tract level. Results showed census tracts with a higher frequency of CR days to have statistically higher rates of poverty and unemployment, along with high proportions of child residents and households without computers. The frequency of CR days and elevated daily PM2.5 concentrations appeared to be strongly related to the occurrence of nearby wildfires, with over 20% of days with sensor-measured average PM2.5 > 35 μg/m3 showing a wildfire within a 100 km radius and over two-thirds of estimated CR days falling on such days with a nearby wildfire. Findings from this study are important to policymakers and government agencies who preside over the allocation of state resources as well as organizations seeking to empower residents and establish climate resilient communities.
        Cover page: Compound Risk of Air Pollution and Heat Days and the Influence of Wildfire by SES across California, 2018-2020: Implications for Environmental Justice in the Context of Climate Change.
        • Article
        • Peer Reviewed
        Faculty Publications (2011)
        Evapotranspiration (ET) is a major pathway for water loss from many ecosystems, and its seasonal variation affects soil moisture and net ecosystem CO2 exchange. We developed an algorithm to estimate ET using a semi-empirical Priestley–Taylor (PT) approach, which can be applied at a range of spatial scales. We estimated regional net radiation (Rnet) at monthly time scales using MODerate resolution Imaging Spectroradiometer (MODIS) albedo and land surface temperature. Good agreement was found between satellite-based estimates of monthly Rnet and field-measured Rnet, with a RMSE of less than 30 W m− 2. An adjustable PT coefficient was parameterized as a function of leaf area index and soil moisture based on observations from 27 AmeriFlux eddy covariance sites. The biome specific optimization using tower-based observations performed well, with a RMSE of 17 W m− 2 and a correlation of 0.90 for predicted monthly latent heat. We implemented the approach within the hydrology module of the CASA biogeochemical model, and used it to estimate ET at a 1 km spatial resolution for the conterminous United States (CONUS). The RMSE of modeled ET was reduced to 21.1 mm mon− 1, compared to 27.1 mm mon− 1 in the original CASA model. The monthly ET rates averaged over the Mississippi River basin were similar to those derived using GRACE satellite measurements and river discharge data. ET varied substantially over the CONUS, with annual mean values of 110 ± 76 mm yr− 1 in deserts, 391 ± 176 mm yr− 1 in savannas and grasslands, and 840 ± 234 mm yr− 1 in broadleaf forests. The PT coefficient was the main driver for the spatial variation of ET in arid areas, whereas Rnet controlled ET when mean annual precipitation was higher than approximately 400 mm yr− 1.
          Cover page: Continental-scale net radiation and evapotranspiration estimated using MODIS satellite observationsCreative Commons 'BY' version 3.0 license
          • Article
          • Peer Reviewed
          UC Irvine Previously Published Works (2021)
          Wildfires can be detrimental to urban and rural communities, causing impacts in the form of psychological stress, direct physical injury, and smoke-related morbidity and mortality. This study examined the area burned by wildfires over the entire state of California from the years 2000 to 2020 in order to quantify and identify whether burned area and fire frequency differed across Census tracts according to socioeconomic indicators over time. Wildfire data were obtained from the California Fire and Resource Assessment Program (FRAP) and National Interagency Fire Center (NIFC), while demographic data were obtained from the American Community Survey. Results showed a doubling in the number of Census tracts that experienced major wildfires and a near doubling in the number of people residing in wildfire-impacted Census tracts, mostly due to an over 23,000 acre/year increase in the area burned by wildfires over the last two decades. Census tracts with a higher fire frequency and burned area had lower proportions of minority groups on average. However, when considering Native American populations, a greater proportion resided in highly impacted Census tracts. Such Census tracts also had higher proportions of older residents. In general, high-impact Census tracts tended to have higher proportions of low-income residents and lower proportions of high-income residents, as well as lower median household incomes and home values. These findings are important to policymakers and state agencies as it relates to environmental justice and the allocation of resources before, during, and after wildfires in the state of California.
            Cover page: Disproportionate Impacts of Wildfires among Elderly and Low-Income Communities in California from 2000-2020.
            • Article
            • Peer Reviewed
            UC Davis Previously Published Works (2023)
            Tracking plant water status is a critical step towards the adaptive precision irrigation management of processing tomatoes, one of the most important specialty crops in California. The photochemical reflectance index (PRI) from proximal sensors and the high-resolution unmanned aerial vehicle (UAV) imagery provide an opportunity to monitor the crop water status efficiently. Based on data from an experimental tomato field with intensive aerial and plant-based measurements, we developed random forest machine learning regression models to estimate tomato stem water potential (ψ stem), (using observations from proximal sensors and 12-band UAV imagery, respectively, along with weather data. The proximal sensor-based model estimation agreed well with the plant ψ stem with R 2 of 0.74 and mean absolute error (MAE) of 0.63 bars. The model included PRI, normalized difference vegetation index, vapor pressure deficit, and air temperature and tracked well with the seasonal dynamics of ψ stem across different plots. A separate model, built with multiple vegetation indices (VIs) from UAV imagery and weather variables, had an R 2 of 0.81 and MAE of 0.67 bars. The plant-level ψ stem maps generated from UAV imagery closely represented the water status differences of plots under different irrigation treatments and also tracked well the temporal change among flights. PRI was found to be the most important VI in both the proximal sensor- and the UAV-based models, providing critical information on tomato plant water status. This study demonstrated that machine learning models can accurately estimate the water status by integrating PRI, other VIs, and weather data, and thus facilitate data-driven irrigation management for processing tomatoes.
              Cover page: Estimation of tomato water status with photochemical reflectance index and machine learning: Assessment from proximal sensors and UAV imagery
              • Article
              • Peer Reviewed
              Faculty Publications (2012)
              We evaluate a new approach to estimate regional evapotranspiration (ET) across a montane, Mediterranean climate gradient in the San Jacinto and Santa Rosa Mountains of Southern California. Spatially distributed evaporative fraction (EF) measurements were made monthly from October 2008 to September 2009 at 54 locations across an elevational gradient using a mobile measurement platform, called the Regional Evaporative Fraction Energy Balance (REFEB) method. We used these measurements and the Enhanced Vegetation Index (EVI) from MODerate resolution Imaging Spectroradiometer (MODIS) observations to derive EF at a regional scale. We converted EF to monthly ET using remote-sensing based observations of available energy. We compared the REFEB ET estimates, along with modified Priestly–Taylor (PT) ET estimates driven by MODIS data against four eddy covariance (EC) towers and eight gauged catchments. Both of the satellite-based ET estimates were highly correlated with tower ET observations (r 2 = 0.66 for REFEB and 0.95 for PT). The PT MODIS approach overestimated ET compared to precipitation estimates and stream gauge measurements, while REFEB ET was moderately lower than PT ET. The annual regional REFEB ET (193 mm) was 87 mm less than precipitation (280 mm). REFEB ET underestimated EC tower ET (regression slope = 0.78, p < 0.001). Regional PT ET (288 mm) exceeded precipitation by 8 mm and significantly overestimated EC tower ET (regression slope = 1.43, p < 0.001). The relationship between precipitation and ET is not linear, with a break around 290 mm/year, at which point ET becomes nearly constant at 200–300 mm/year with increasing precipitation. This causes a break in the precipitation–runoff relationship, with a disproportionate increase in runoff when precipitation exceeds 290 mm/year. REFEB provides a viable method to estimate regional ET, which is applicable to areas that are poorly constrained by other remote sensing approaches.
                Cover page: Assessing regional evapotranspiration and water balance across a Mediterranean montane climate gradientCreative Commons 'BY' version 3.0 license
                • Article
                • Peer Reviewed
                UC Irvine Previously Published Works (2014)
                Wildfire ignition requires a combination of an open spark, and suitable weather and fuel conditions. Models of fire occurrence and burned area provide a good understanding of the physical and climatic factors that constrain and promote fire spread and recurrence, but information on how humans influence ignition patterns is still lacking at a scale compatible with integrated fire management. We investigated the relative importance of the physical, climatic and human factors regulating ignition probability across Southern California's National Forests. A 30-year exploratory analysis of one-way relationships indicated that distance to a road, distance to housing and topographic slope were the major determinants of ignition frequency. We used logistic and Poisson regression analyses to model ignition occurrence and frequency as a function of the dominant covariates. The resulting models explained ∼70% of the spatial variability in ignition likelihood and 45% of the variability in ignition frequency. In turn, predicted ignition probability contributed to some of the spatial variability in burned area, particularly for summer fires. These models may enable estimates of fire ignition risk for the broader domain of Southern California and how this risk may change with future population and housing development. Our spatially explicit predictions may also be useful for strategic fire management in the region.
                  Cover page: Controls on the spatial pattern of wildfire ignitions in Southern CaliforniaCreative Commons 'BY' version 4.0 license
                  Top