Scientific Investigations Report 2006–5043
U.S. GEOLOGICAL SURVEY
Scientific Investigations Report 2006–5043
Daily 
were computed from daily Bowen-ratio ET and daily 
for three relative densities of phreatophytes. A modified curve was derived empirically from average daily. Daily modified were developed for barren soils. Daily and monthly ET was recomputed using the modified daily and daily .
Modified coefficients can be computed for each ET station by dividing daily energy-budget ET by daily LCRAS . A curve can be developed from the computed daily coefficients by averaging the daily coefficients for two periods of the year that correspond to growth stages of plants (generally, dormancy during the cool season and vigorous growth during the summer) and computing sloped, straight lines between the end and start of each period (fig. 11).
ET coefficients represent average plant and soil conditions (growth stage, stress, or density or soil salinity) over space and time and may not accurately reflect the conditions of any particular vegetated area on a specific day. Given that coefficients represent average conditions, the accuracy of the modified coefficients may be improved by deriving average coefficients from data for several years, rather than 1 year, thereby reducing the effects of large daily fluctuations of reference ET and seasonal variability. For example, large daily fluctuations of the computed coefficients for the SC station for 2003 (fig. 11) are due to corresponding large daily fluctuations of reference ET (fig. 7). Unexpected seasonal variability of computed daily coefficients, such as low values in September 2003 followed by high values from October to November 2003, probably are due to anomalous reference ET.
Available daily Bowen-ratio ET data from 2002 through 2004 were divided by corresponding daily to develop an average daily ET coefficient curve for each station. Daily coefficients were averaged for two separate periods (late autumn or early winter to early or late spring, and early summer to early or late autumn) to empirically derive a modified daily curve. Periods were based on how an average value changed during a period. A linear equation was used to interpolate daily coefficients for the two transition periods.
Daily were computed for the SC station from May 23, 2002, to June 14, 2004. An average modified daily coefficient (fig. 12A, table 9 at back of report) was derived from available computed coefficients for each day of the year and is the average of 2 or 3 days, based on when data were available. The modified daily for the SC station ranges from 0.22 (January 1–March 10 and November 30– December 31) to 0.76 (May 19–September 25). In contrast, the LCRAS daily from May to September is 1.10 for sc_high group (fig. 12A).
Daily ET coefficients were computed for the MV station, May 23, 2002–June 14, 2004. An average modified daily coefficient (fig. 12B, table 10 at back of report) was derived from available computed coefficients for each day of the year and is the average of 2 or 3 days, based on when data were available. The modified for the MV station ranges from 0.30 (January 1–February 21 and October 26–December 31) to 0.53 (May 9–July 23). In contrast, the LCRAS daily from May–September is 1.16 for sc/ms/aw group and 0.97 for ms/aw group (fig. 12B).
Daily ET coefficients were computed for the AW station from January 1, 2003, to June 14, 2004. An average modified daily coefficient (fig. 12C, table 11 at back of report) was derived from an average of variable number of days from January 1 to June 14. From June 15 to December 31, the derived coefficient is equal to the computed daily coefficient for 2003. The modified for the AW station ranges from 0.21 (January 1–February 21 and November 26–December 31) to 0.56 (May 7–August 17). In contrast, the LCRAS daily from May–September is 0.86 for low_veg group and 0.99 for aw group (fig. 12C).
Due to the lack of Bowen-ratio ET data, modified coefficients could not be empirically derived, though it is possible to evaluate the LCRAS and 
for the barren group. The LCRAS curve for barren areas was modified (fig. 13, table 12 at back of report) using representative ET rates estimated for other barren-soil, low-density vegetation, or sparse-vegetation areas in southern Nevada (table 1). While developing the modified LCRAS curve for “dry” barren areas, it was assumed that annual ET is less than 1 ft. Also, certain criteria were established: (1) the modified barren soil coefficient curve retains the original LCRAS coefficient from mid-April to mid-October, (2) the modified coefficients are lower than the LCRAS coefficients from January through March, and (3) the modified coefficients are the same for December and January.
Daily and monthly ET for the SC station were computed and compared to ET estimated by the Bowen-ratio method. Daily ET was computed from May 23, 2002, to June 14, 2004, using the modified daily and daily and compared to the daily Bowen-ratio ET for the same period. Data compared for 754 days had a correlation coefficient of 0.92, with large differences for many days (fig. 14A). Monthly ET was computed (sum of daily ET) and compared to Bowen-ratio ET for 24 months (June 2002–May 2004; fig. 14B). Based on a correlation coefficient of 0.99, monthly ET computed using a modified daily yields results that compare favorably with monthly Bowen-ratio ET. Total ET for two 12-month periods (June 2002–May 2003 and June 2003–May 2004; table 6) also was compared to Bowen-ratio ET; computed ET was about 3.7 and 2.6 percent more, respectively, than the Bowen‑ratio ET.
Daily and monthly ET for the MV station were computed and compared to ET estimated by the Bowen-ratio method. Daily ET was computed from May 23, 2002, to June 14, 2004, using the modified daily and daily and compared to the daily Bowen-ratio ET for the same period. Data compared for 754 days had a correlation coefficient of 0.89, with large differences for many days (fig. 15A). Monthly ET was computed and compared to Bowen-ratio ET for 23 months (June 2002–May 2004, March 2004 was omitted due to poor energy-budget data; fig. 15B). Monthly percentage of differences ranged from ‑77.1 to 23.2 percent (November 2002 and October 2002, respectively), with an average difference of 6.7 percent. The magnitude of differences ranged from -0.07 to 0.04 ft (August 2003 and October 2002, respectively), and averaged 0.01 ft. Based on a correlation coefficient of 0.97, monthly ET computed using a modified daily yields results that compare favorably with monthly Bowen-ratio ET.
Total ET for one 12-month period (June 2002– May 2003) and one 11-month period (June 2003–May 2004, excludes March 2004) were 0 and 8.6 percent more, respectively, than the total ET estimated by the Bowen-ratio method (table 6).
Daily and monthly ET for the AW station were computed and compared to ET estimated by the Bowen-ratio method. Daily ET was computed from January 1, 2003, to June 14, 2004, using the modified daily and daily and compared to the daily Bowen-ratio ET for the same period. Data compared for 531 days had a correlation coefficient of 0.89, with large differences for many days (fig. 16A). Monthly ET was computed and compared to Bowen-ratio ET for 17 months from January 2003 to May 2004 (fig. 16B). Monthly differences ranged from ‑66.9 to 21.2 percent (January 2004 and November 2003, respectively), with an average difference of 13.3 percent. The magnitude of differences ranged from -0.11 to 0.02 ft (September 2003 and May 2004, respectively), with an average of 0.02 ft. Based on a correlation coefficient of 0.98, computed monthly ET compare favorably with monthly Bowen-ratio ET.
Total ET for one 12-month period (January–December 2003) and one 5-month period (January–May 2004) also were computed and equaled 12.3 percent and 1.0 percent more, respectively, than estimated Bowen-ratio ET (table 6).
The modified barren-soil ET coefficients were used to compute monthly ET (fig. 17). By applying the modified coefficients, estimated annual ET for barren-soil areas was reduced to less than 1.0 ft, with cool-season monthly ET lower than warm-season ET. For 2002, estimated ET equaled 0.96 ft and for 2003 and 2004, annual ET was estimated at 0.87 and 0.88 ft, respectively.
The magnitude and temporal distribution of modified ET coefficients are a function of vegetation density, type and composition, and depth to ground water. Generally, the coefficient increases with vegetation density, which varies throughout the year, and decreases with increasing depth to water. The type of vegetation affects the seasonal distribution of ET. Deciduous plants, such as saltcedar and mesquite, have high coefficients during the warm season, but low coefficients during the cool season when they are dormant (fig. 18). Evergreen plants, such as salt grass, baccharis, and arrowweed will have higher coefficients during the cool season than deciduous plants because they are not completely dormant. An area of homogeneous vegetation probably will have a shorter transition period, because the plants ‘green-up’ and go dormant at about the same rate; in contrast, a mixed-plant community probably has a longer transition period, because the plants green-up at different rates.
The relatively high modified coefficient during the warm season for the SC station reflects high-density vegetation (fig. 18). The relatively low modified coefficient during the cool season reflects complete dormancy of vegetation. Because the SC station consists of homogeneous saltcedar, all the plants are dormant during the cool season and loss of water is restricted mostly to evaporation from the soil. The transition period is relatively short due to homogeneity of the vegetation.
The mid-range modified coefficient for the MV station reflects medium-density vegetation and, as a result, has a lower coefficient during the warm season (fig. 18). However, during the cool season, coefficients for the MV station are higher than for the SC station due to the presence of evergreen plants, such as salt grass, baccharis, arrowweed, and sharp-leaved juncus. Vegetation at the AW station is low-to-medium relative density and, as a result, has about the same modified coefficient during the warm season as the MV station.
To further evaluate the modified coefficients, monthly ET computed from the modified coefficient and reference ET method for each water year (October–September) were compared (figs. 19A–19C). Annual variability of monthly ET may be due to differences in actual ET, which are reflected by differences in , inherent error in the energy-budget method (typically estimated to be about 5 percent), and averaging of daily modified coefficients over time.
For more information about USGS activities in Nevada, visit the USGS Nevada Water Science Center home page .
U.S. Department of the Interior |
U.S. Geological Survey
Persistent URL: https://pubs.water.usgs.gov/sir20065043
Page Contact Information: Publications Team
Page Last Modified: Thursday, 01-Dec-2016 18:55:55 EST
