2014-05-15_REVISION - M1987064 - Laserfiche WebLink

Home Browse Search

296 D.C. Goodrich et al. /Agricultural and Forest Meteorology 105 (2000) 281 -309 presented in individual subsections. Section 5.1 per- tains to mesquite ET and Section 5.2 pertains to results regarding the C/W transpiration measurements, their scaling and calibration of the P —M model. Section 5.3 contains the results of the water balance compu- tations at both the stand and corridor level. Finally, in Section 5.4, results and discussion regarding the scaling to these riparian ET estimates to the overall corridor are presented. The first brief result, from Scott et al. (2000) re- garding the sacaton grass, resulted in a simplification of the approach employed herein. While the saca- ton grasslands were originally envisioned to utilize groundwater, Scott et al. (2000) concluded that ET from most of the sacaton in the Lewis Springs por- tion of the San Pedro originates almost entirely from near -term precipitation and from soil moisture stor- age. They noted that the 1997 yearly precipitation at the sacaton site was 247 mm with an increase of 8 mm of soil moisture, while the total evaporation loss measured at the sacaton tower was 272 mm (see Figs. 3 and 4a of Scott et al., 2000). Assuming a 20% error in the Bowen ratio and soil moisture measure- ments, and a 10% in the rainfall measurements, it was assumed that all evaporative losses from the sacaton originated from rainfall and soil moisture changes. Therefore, groundwater extraction via transpiration from the sacaton grass was not considered significant, and was not considered as part of the water balance computations for the entire corridor. The second result regarding the remotely sensed thermal data also simplified the approach to the overall riparian corridor ET estimation. It was hypothesized that the riparian vegetation in the well- watered peren- nial reaches would exhibit significantly cooler surface temperatures due to greater transpiration than more highly stressed vegetation in intermittent reaches of the river. If true, this would justify using the remotely sensed surface temperature data to spatially parti- tion the riparian corridor to differentially estimate ET when reaches of the river became intermittent. To test this hypothesis, remotely sensed surface tem- peratures of several well- defined cottonwood clusters at nearly identical times near Lewis Springs and the Boquillas Ranch were examined for the April, July, and August flight dates. The limited data indicated a relatively small increase in the temperature difference (0.7- 1.2 °C) between the dry Boquillas reach and the flowing Lewis Springs reach for the July overflight. Even though ground observations noted stressed veg- etation and a loss of leaves at the dry intermittent site, the hypothesized increase in surface temperature may have been masked by understory conditions or the resolution of the sensor. Partitioning the riparian corridor for differential computation of riparian ET based on the acquired thermal remote sensing data was therefore not justified. 5.1. Mesquite ET To scale the Bowen ratio measurements from the mesquite tower using the classified land cover area, it was necessary to derive mesquite ET estimates for the equivalent of 100% mesquite cover. To accomplish this, the following simple linear partitioning of the fluxes at the mesquite tower was assumed: BRm = (ETm + Im)Am + (ETssb + Issb)Assb (11) where BRm is the measured Bowen ratio flux at the mesquite tower, ETm the mesquite ET, Im the amount of intercepted rainfall contributing to the fluxes mea- sured at the Bowen ratio tower, Am the percentage area of mesquite contributing to the fluxes measured at the Bowen ratio tower, and ETssb, Issb, and Assb are comparable quantities for the sacaton, scrub and bare areas contributing to the fluxes at the mesquite Bowen ratio tower. Scott et al. (2000) estimated the percentage area of mesquites contributing to the measured fluxes at the mesquite Bowen ratio tower was approximately 50 %, while sacaton, scrub and bare areas comprised the other 50 %. For this analysis, is was assumed that the quantity (ETssb + Issb) was equal to the evapora- tive fluxes measured at the sacaton Bowen ratio tower as these three cover classes comprised virtually the entire area contributing to the fluxes measured at that tower. Interception was assumed to equal to first 3 mm of rainfall of any event based on measurements by Tromble (1983) on desert tarbush. Given the large un- certainties in defining the source areas for flux mea- surements, both of the percentage area estimates in Eq. (11) were assigned errors of 50 %. The estimates obtained for mesquite ET for the water balance period (DOY 101 -191) and for the entire growing season (DOY 101 -294) were 184 and