2014-05-15_REVISION - M1987064 - Laserfiche WebLink

Home Browse Search

D. C. Goodrich et al. /Agricultural and Forest Meteorology 105 (2000) 281 -309 402 mm, respectively. Partitioning of mesquite ET derived from surface water or groundwater sources was not possible with the measurements made. The entire - scaled mesquite ET estimate for both the wa- ter balance and the growing season was assumed to be derived solely from groundwater. This appears to be a reasonable assumption during the pre- monsoon water balance period, but limited measurements pre- sented in Snyder and Williams (2000) indicate that the mesquite can readily use surface water when it becomes available. During the monsoon, after sig- nificant rains and runoff events occurred, several mesquites that Snyder and Williams (2000) measured, were able to derive over 50% of their transpired water from surface sources. Therefore, the growing season estimate of the amount of groundwater used by mesquites obtained from the mesquite tower is conservatively large (i.e. if the mesquites are able to use surface water during the monsoon then less groundwater from the regional aquifer will be used). Clearly more detailed isotopic and flux measurements (given adequate time and resources) must be made on a variety of mesquites to fully understand their water sources and transpiration quantities. 5.2. C/W transpiration and P —M model calibration Schaeffer et al. (2000) present the sap flux measure- ments and results for estimating C/W transpiration at the trees and patch scale. The results discussed herein focus on scaling these measurements and our ability to model them. In examining the patch level results, it was found that the variance in sapwood area to canopy area was relatively large across the patches. This was in part attributed to the differences in canopy struc- ture between the young or newly established patches and the older patches. Schaeffer et al. (2000) con- cluded that the differences between the young and old patches in water use per unit canopy area were sig- nificantly different. However, at the patch scale, good agreement existed between the sap flow estimates of C/W transpiration and independently derived esti- mates of ET from the scanning Raman LIDAR. Both Cooper et al. (2000) and Eichinger et al. (2000) found an RMSE between the sap flow and LIDAR -based ET estimates of approximately 0.03 mm h -1 or 0.36 mm per day assuming a nominal 12 h of transpiration per day. Another factor, noted above, which makes 297 comparison across the patches problematic is the un- certainty in estimating the area of the patch. For the relatively small patches (444 -1985 m2), the ratio of patch perimeter to area is relatively large. This uncer- tainty could also have contributed to the high variance in sapwood area to canopy area across the patches. As noted in Section 4.1, we assumed that sampling a larger number of trees to scale from the patch to stand level provided a more certain estimate of the canopy area as well as a more representative sample of trees. The sap flux estimates of C/W transpiration scaled to the stand level on a per unit canopy area basis for DOY 158 and 159 of the 2238 trees at Lewis Springs are illustrated in left -hand portion of Fig. 5. The er- ror bands for these estimates computed using Eq. (7) are also included in this portion of the figure. These error estimates include the errors associated with the sapwood area versus tree diameter regression and the standard error of the sap flux measurements. Also in- cluded in the right -hand side of this figure is a com- parison between stand level C/W transpiration at the well- watered Lewis Springs site and the ephemeral Escapule Wash site for DOY 192. The fluxes for the stressed Escapule site are less than half of those at Lewis Springs. However, it should be noted that these trees are in a side wash approximately 250 -300 m away from, and substantially higher than, the main channel of the San Pedro. These trees have less access to groundwater than those near the main channel and lower in the floodplain. This situation is not common in the reach between Lewis Springs and Charleston but it does demonstrate the large variation of sap flux depending on the trees access to groundwater. The scaled stand level estimates of C/W transpira- tion at Lewis Springs for each SMP (e.g. the left -hand side of Fig. 5) were used to compute the bulk canopy level resistance using Eq. (5) as described in Section 4.2.1. This procedure resulted in an average daily cal- ibrated bulk canopy resistance for each SNIP. A con- stant daily average value of canopy resistance was used in subsequent computations even though stom- atal resistance (and hence canopy resistance) is known to fluctuate diurnally. For each multi -day SMP, an av- erage canopy resistance was obtained by averaging the daily calibrated resistances for each day of the SNIP. For example, the average daily resistances com- puted for DOY 191, 192, 193, and 194 were 123, 144, 192, and 182 s m -1, respectively. The average value of