1061
ESTIMATING POTENTIAL AND ACTUAL DAILY TRANSPIRATION OF
CROP CANOPIES FROM REMOTELY SENSED SPECTRAL DATA AND
BASIC METEOROLOGICAL DATA - THEORY AND INITIAL TEST -
Y oshio I NOUE*. M. Susan MORAN**, and Paul J. PINTER JR.**
* National Institute of Agro-Environmental Sciences, Tsukuba, Iharaki, 305 Japan
**U.S. Water Conservation Laboratory, 4331 E. Broadway Rd., Phoenix, AZ, 85040 USA
ABSTRACT:
It is essential to estimate transpiration separately from évapotranspiration for better understanding of water and
energy exchange processes in the soil-plant-continuum and for more efficient water use. Our objective is to
estimate the potential and actual values of daily transpiration rate in vegetation canopies. A new concept for
estimating potential and actual values of daily transpiration rate of vegetation canopies is presented along with
results of an initial test. The method is based on a physical foundation of spectral radiation balance for a
vegetation canopy, the key inputs to the model bang the remotely sensed spectral reflectance and the surface
temperature of the plant canopy. The radjation interception or absorptance is estimated more directly from
remotely sensed spectral data than it is from the leaf area index. The potential daily transpiration is defined as a
linear function of the absorbed solar radiation, which can be estimated using a linear relationship between the
fraction absorptance of solar radiation and the remotely sensed Soil Adjusted Vegetation Index for the canopy.
The actual daily transpiration rate is estimated by combining this concept with the Jackson-Idso Crop Water
Stress Index, which also can be calculated from remotely sensed plant leaf temperatures measured by infrared
thermometry. An initial demonstration with data sets from an alfalfa crop and a rangeland suggests that the
method may give reasonable estimates of potential and actual values of daily transpiration rate over diverse
vegetation area based on simple remote sensing measurements and basic meteorological parameters.
KEYWORDS: Canopy Temperature, Infrared Thermometry, Remote Sensing, Spectral Reflectance,
Transpiration, Vegetation Index, Water Stress.
1. INTRODUCTION
Biomass production and yield of plants have been correlated to the amount of canopy water use ( Hanks 1983,
Howell 1990). In fact, both C0 2 and HjO exchange processes between plant leaves and the atmosphere are
primarily regulated by stomata] aperture. Thus, effective water management is one of the most important issues
in agricultural production, especially in arid and semi-arid regions. Efficient water use at the regional scale
will become more and more important because of increasing impact of global climate change and population
growth. Remote sensing techniques will be the effective method for estimating water consumption over wide
areas of crop canopies as well as natural vegetation.
Evapotranspiration (ET), i.e. the total value of evaporation from die soil surface and transpiration from
plant leaves, has long been estimated with lysimeter measurements and by meteorological methods. However, it
seems to be essential to estimate transpiration separately from évapotranspiration for better understanding of
water and energy exchange processes in the soil-plant-atmosphere continuum as well as for more efficient water
use in plant production. As for the transpiration alone, heat pulse method and heat balance method have been
applied to the estimation of sap flow rates of indivuidual plants (Cohen et al. 1993). Neverthless, they both
need to attach gauges to the stems of each plant Few methods are available for estimating the transpiration rate
(Tr) of intact plants over wide areas although remotely sensed leaf or canopy temperatures are useful for
detecting stress-response and for estimating the transpiration of single leaf (Inoue et al. 1990).
Our objective, thus, is to estimate the energy-limited potential values and actual values of daily transpiration
rate (Tr p and Tr a ) in vegetation canopies. Daily estimates of these parameters seem to be sufficient for such
applications as crop diagnosis, yield prediction, management decision-making, and ecological plant science. In
this paper, we present the new concept along with the results of pre liminar y tests. The method is based on a
physical foundation of spectral radiation balance for the vegetation, the key inputs to the method being the
remotely sensed spectral vegetation index and the surface temperature of plants.
