Measuring Evapotranspiration in Urban Irrigated Lawns.

Water conservation is becoming increasingly critical in metropolitan areas.  The use of automated lawn-irrigation systems is rising and maladjusted systems may apply excessive water.  Micrometeorological methods were used to estimate actual lawn-water use (evapotranspiration; ET) in urban lawns.  Small weather stations (tripods) were deployed in residential yards in Manhattan, KS, USA, to estimate reference ET (ET_o) via FAO-56.  Actual evapotranspiration (ET_actual) was measured within five km from well-watered turfgrass using eddy covariance (EC); a tripod was deployed in the EC source area to provide reference ET at the site (ET_o,_{EC tripod}).  EC data were used to compute “lawn coefficients” (K_c) for individual lawns by determining ratios of ET_o from lawn-deployed tripods to ET_o from the tripod at the EC site (ET_{o,EC tripod}); hence, K_c = ET_o,tripod / ET_o,_{EC tripod}.  Individual lawn coefficients ranged from 0.49 – 0.79 with the exception of one outlier (0.21). Multiplying individual K_c,i by ET_actual,EC yielded ET_actual for individual yards.  By averaging all lawn-specific coefficients (K_c,i), the entire area was represented by one lawn coefficient (K_c) of 0.639.  Estimates of ET_actual for the area were determined by multiplying 0.639 by ET_actual,EC.  Cumulative ET_actual,EC at the EC site for the period corresponding to tripod deployments in residential lawns averaged 10.9 ± 1.9 mm, while cumulative ET_actual in residential lawns using K_c = 0.639 averaged 6.9 ± 0.8 mm; lower windspeeds and shaded areas in lawns contributed to lower ET compared with the EC site.  Estimates of average cumulative ET_actual using K_c,i from individual lawns was 6.2 ± 1.8 mm.  Determining ET_actual for individual lawns using lawn-specific coefficients (K_c,i) was to within ±10.9% of ET_actual calculated using one overall K_c.  Considering differences in microclimates between the EC site and individual lawns, estimates of in-town ET found by correcting EC data using one average lawn coefficient were good.