Remote Sensing of Urban Oasis Actual Evapotranspiration to Aid Irrigated Landscape Water Conservation

Water conservation is urgent for irrigated urban landscapes in much of western and southeastern U.S. as changing climate amplifies high temperature and drought extremes. Estimated evapo-transpiration (ET) is the key tool in conservation programs in drought afflicted cities to guide precision irrigation timing and duration to meet plant water needs, and set targeted water alloca-tions for end user compliance. As adopted from agriculture water management, urban ET is es-timated in a two part process. First, reference evapotranspiration (ETo) is calculated (Penman-Monteith energy balance equation) for a hypothetical, standardized turfgrass surface from weath-er data (solar radiation, wind speed, air temperature, vapor deficit) measured over a large, well irrigated urban turf surface such as a park or golf course. Then, actual landscape plant ET is es-timated as ETo times an empirically-derived ratio factor specific to a plant type. Unlike agriculture, these ratio factors are specific to very broad plant types, turfgrass and woody plants, rather than a particular crop species. These factors (plant factors, or PF) simplify enor-mous biological complexity in water use traits among the many landscape species to two robust water use categories distinct each for turf and woody plants. Turfgrass separates into two PF's based on photosynthate metabolism, C4 (PF=0.6) and C3 (PF=0.8). Woody plants separate into two PF's based on transpiration (stomatal) sensitivity to dry air (vapor deficit) based on climate: humid (PF=0.7) and arid (PF=0.5). The American Society for Agricultural and Biological Engi-neers is developing a model standard (Standard-623) for codifying these plant factors in irrigated urban landscapes.

This urban ETo approach has two problems: ETo-PF poorly capture urban heat effects on plant water use and so precision landscape irrigation is at best approximate; ETo is a measure of the past and so poorly fits anticipated water allocations. To address these two issues, we are investi-gating satellite (daily MODIS downscaled to weekly Landsat) thermal-multispectral imagery to estimate actual urban turfgrass ET to replace ETo-PF. Actual ET can be extracted from the rela-tionship between cool (parks, golf courses) to hot (pavement, parking lots) pixels and vegetation cover (visible spectral bands), and so better captures urban heat island effects. We are also in-vestigating reconstruction of paleo ET over the past 500 years in northern Utah from ring width of low elevation junipers whose growth is strongly controlled by the same stomatal response to dry air governing the arid PF=0.5 for woody plants. Paleo ET may be linked to regional wet/dry oscillations driven by mid-Pacific sea surface temperatures, and so amenable to modeling for better year-to-year tailoring of end user water allocations.