Trends in Vegetation Greenness-based Evapotranspiration Rates for Riparian Corridors
Pamela Nagler1*, Ibrahima Sall2, Armando Barreto-Muñoz3, and Kamel Didan3
1U.S. Geological Survey, Southwest Biological Science Center, 520 North Park Avenue, Tucson, AZ 85719 USA
2Department of Agricultural and Resource Economics, University of Arizona Tucson, AZ 85721, USA
3Biosystems Engineering, University of Arizona, Tucson, AZ, 85721 USA
Estimating riparian vegetation extent, magnitude of greenness, changes in phenology and plant water use requires novel science methods for monitoring changes as accurately as possible with remote sensing tools. In these narrow riparian corridors, we quantify the loss of water from leaves and soil as one variable, actual evapotranspiration (ETa). Increases in ETa are indicative of increasing vegetation cover and therefore, increasing ‘losses’ of water through ETa represent positive trends in riparian ecosystem health; decreasing ETa may indicate dwindling riparian cover due to less available water for canopy growth due to drought, groundwater flux, beetle defoliation, fire, or increasing salinity. The objectives of this study were to calculate annual ETa (mmyr-1) for selected riparian areas in the Lower Colorado River Basin, mostly in Arizona. The riparian areas were delineated and monitored using the two-band Enhanced Vegetation Index (EVI2). We acquired 30-m resolution Landsat scenes, processed and performed a pixel-wise quality assessment to remove pixels with high aerosols and clouds, and computed EVI2 every 16-days over 20 years. We then computed daily potential ET using the Blaney-Criddle formula with input temperature data from gridded weather data using Daymet (1 km). Riparian ETa was quantified using the Nagler ET(EVI2) model to produce time-series data for two different periods within the years 2000-2022.
Most rivers in these drylands have regulated flows, such as the Lower Colorado River mainstem. We present results from the Salt, Gila and Santa Cruz Rivers, Bill Williams and San Juan, as well as unregulated rivers (e.g., San Pedro, Virgin). A biocontrol beetle (Diorhabda spp.) defoliated much of the saltcedar (Tamarix spp.) riparian corridor landcover in these Arizona rivers starting in ca. 2008; for example, on the Lower Colorado River mainstem, monitoring of riparian ET(EVI2) showed annual totals with large declines believed to be in part due to biocontrol. However, over the last decade (2013-2021), areas have been restored following the implementation of the Lower Colorado River Multi-Species Conservation Plan. In the Colorado River delta, the average of four restoration sites, which cover only ca. 3% of the riparian corridor showed ET(EVI2) increased by 22.1% (0.60 mmd-1) post-planting (since ca. 2014). From 2000 to 2021 or 2022, various rivers were studied to determine the average annual ETa (mmyr-1) for riparian corridors, unrestored areas, and restored areas. Note that the period of study varies by processing dates; however, our study findings for riparian ETa are as follows: the Bill Williams River ETa decreased from 925 to 600 mmyr-1 (2014-2022); the San Pedro River ETa increased from 975 to 1075 mmyr-1 (2014-2022); the Virgin River ETa increased from 675 to 825 mmyr-1 (2014-2022); the Salt River ETa increased 13.7% from 800 mmyr-1 to 910 mmyr-1 (2000-2021); the Gila River ETa increased 2.7% from 725 mmyr-1 to 745 mmyr-1 (2000-2021), with occasional periods of decreases (e.g., 2002, 2013) followed by increases; and the Santa Cruz River ETa had the most significant increase of 24.0% from 770 mmyr-1 to 955 mmyr-1 (2000-2021). The increasing trends on these rivers could be due to riparian species composition altered by the tamarisk beetle followed by secondary or replacement species which established green canopies, restoration efforts or other changes in water or land management. This study provides valuable estimates of riparian water use that may assist with decision-making by natural resource managers tasked with allocating water and managing habitat along these riparian corridors.