3.2 Summer Energy and Water Flux Comparisons
Most of the annual ET occurs over the summer months (Table 3), and closer errors are lowest during these times, so we then compared datasets of summers (June-August) 2017, 2018, and 2019. We averaged latent heat, total net radiation (Rn-G), air temperature, and dewpoint depression for each summer using the same dates for all variables across each summer. We calculated ET (latent heat flux multiplied by the latent heat of vaporization) for all three summers along with the total ET for the water year (water year 2017 is shortened and begins when we installed the tower in April 2017). This table includes the total precipitation for each water year, which was obtained from a precipitation gage located within the East River basin at an elevation approximately 200m higher and 6.2km northwest of the flux tower (Billy Barr station). Average latent heat flux is greatest during the summer 2018 with summer 2019 having the lowest average latent heat flux (Table 3). The table shows that the differences between summers of available energy, ET, latent heat, and air temperature are much less across the three summers than the differences in precipitation. 2019 had the greatest precipitation with the year of lowest precipitation, 2018, being 400mm less. Total summer ET does not follow the same patterns with the largest ET value occurring during the lowest precipitation year. Similar to the findings of Scott et al. (2008), ET is not as variable as precipitation over the three study years. While precipitation varies by 30% across the three years, ET varies by only 14% suggesting that precipitation is not the primary driver (or limiter) of ET in this system and confirming that this site is generally a saturated end-member site, which is not water-limited for ET.
We then compared these summers on a time series (Figure 6). All three summers have similar monthly patterns and magnitudes of ET showing overall site consistency across three water years despite large differences in precipitation (Figure 6b). However, ET is at times inversely correlated to soil moisture at both 5cm and 25cm (Figure 6c,d). For example, figure 6d shows soil moisture at 5cm and we see that May and June respond to snowmelt with the highest average soil moisture in 2019. However, ET for May and June of 2019 is the lowest of the three summers. Available energy also does not correlate perfectly with the patterns seen in ET across the three years (Figure 6a). The years with greatest average monthly available energy do not always result in years with the greatest average monthly ET. This data, along with the data shown in table 3, seems to suggest that ET and soil moisture are, at times, inversely correlated indicating that water for ET comes not just from shallow soil moisture, but also from other sources such as deeper groundwater or ponded water.
The drying of soil moisture throughout the summer can also be observed in the May through September groundwater use values (Table 4). This is a critical time for vegetation as the spring snowmelt recharges soil moisture and groundwater and this plant water source is drained by vegetation until the availability of summer rainfall. There is substantial evidence that plants also access groundwater, particularly during this time in mountain systems (Bearup, Maxwell, Clow, & McCray, 2014). Using an equation from Scott et al. (2008) (Equation 3), we can estimate how much ET comes from groundwater over the summer for years 2017, 2018, and 2019 as follows:
\(\text{ET}_{\text{gw}}=ET-(P-S)\), Eq. 3
where ETgw is ET from groundwater, ET is total evapotranspiration, P is precipitation, and ∆S is the change in soil moisture in the top 30cm of soil from May to September. Runoff is assumed to be negligible since we are calculating ETgw over the area of the eddy flux tower footprint. Positive ETgw indicates ET is greater than precipitation and soil moisture change resulting in ET drawing from older groundwater. The ETgw values for all three summers are positive indicating that groundwater supplies a fraction of ET regardless of precipitation. ET drew most heavily from groundwater in the summer of 2017 (76.2%) closely followed by 2019 (75.5%), with the summer of 2018 having the least amount of groundwater use. While 2018 had the least precipitation annually (Table 3), much of the annual precipitation occurred during the summer months leading to less groundwater use than the other two summers showing a reliance on rainwater for summer 2018 rather than a reliance on groundwater due to snowmelt. These summers offer insight into the variability of ET groundwater use across water years. Though the magnitude of precipitation is crucial for ET, the timing of precipitation across the year dictates whether ET needs to draw from groundwater as in 2017 and 2019 when ET seems to use water from snowmelt, or whether ET coincides with summer rain as in the case of 2018. This shift in water sources is suggested by simulations (Bearup et al., 2014; Kollet & Maxwell, 2008; Maxwell & Condon, 2016; Maxwell, Condon, Danesh-Yazdi, & Bearup, 2019) and corroborated by shallow groundwater observations on a hill slope adjacent to the tower site that confirm that vegetation would have access to groundwater along riparian flood plain (Tokunaga et al., 2019).
Without access to groundwater, ET values would decrease substantially in 2017 and 2019 as the only contributing water, we estimate, would be from precipitation and soil moisture resulting in 71.82mm and 65.21mm of ET for summers of 2017 and 2019, respectively. These ET values are what we might expect as a low end-member at higher elevations on ridge tops where land-surface energy processes may be disconnected from groundwater (Kollet & Maxwell, 2008; Maxwell & Kollet, 2008), and may not have as much access to groundwater as the flux tower location, which sits in a convergent zone. While subject to uncertainty, these estimates indicate that groundwater may increase ET values by up to 76% making it critical to better constrain these higher elevation water fluxes. As soil structure and soil moisture conditions vary across the East River basin it is important to provide additional observations of ET to constrain this variability.