17 Irrigation and Drainage
383
of images gives then a time series I(t) comprising the values of the selected
spectral index on all selected dates. The I(t)-values give a measure of crop
phenology and can be used for crop identification. To this end images may be
processed to enhance contrast, so that qualitative interpretation and delineation is easier, or numerical classification procedures. may be applied to the
I(x,y,t) images.
C. Calculation and mapping of crop water requirements. Traditionally (F AO,
1977) crop water requirements are estimated by combining climate data with crop
specific coefficients, which accounts for the ratio of the maximum crop specific
evapotranspiration to a reference evapotranspiration dependent on climate only.
More recently (e.g. Allen et aI., 1994) FAO guidelines are evolving towards the
use of the Penman- Monteith (PM) equation which avoids the use of empirical
crop coefficients. The rationale for this approach is that empirical values of crop
coefficients, such as provided by FAO (1977), cannot account for the actual
variability of canopy properties which control consumptive water use. The two
approaches are less different than it may be concluded at a preliminary analysis. It
can be shown, using the PM equation, that the crop coefficients are related
explicitly to the parameters of the PM equation (Stanghellini et aI., 1990).
Accordingly, radiance measurements may contribute in different ways to the
determination and mapping of crop water requirements:
accurate determination of cropped area (see A) reduces significantly the
uncertainty on crop water requirements within an irrigation scheme.
mapping of specific crops (see B) in combination with accurate knowledge of
crop phenology reduces further the uncertainty on the spatial and temporal
variability of crop water requirements.
clusters of crops having similar crop coefficients can be mapped using
advanced numeric classification procedures (D'Urso and Menenti, 1996a, b);
estimates of surface reflectance, leaf area index, surface temperature can be
derived from spectral and directional radiances and applied to compute maps
of crop water requirements.
estimates of net radiation, Rn, and soil heat flux, G, can be obtained from
spectral radiances, with (Rn - G) being an useful upper bound of maximum
evapotranspiration and, therefore, of crop water requirements.
D. Crop stress detection using visible through near infrared radiances.
Spectral reflectance of leaves is very sensitive to water absorption, with the effect
being observable even in the broad bands of Thematic Mapper (Jackson et aI.,
1980). At higher spectral resolutions (e.g. 10 nm), water absorption is observable
at multiple wavelengths (Curran, 1989). Determination ofleafwater content gives
a direct measure of water shortage. Relationships between properties of green
foliage, e.g. leaf water content, and reflectance spectra of canopies change (e.g.
Wessmann, 1992) with canopy architecture and background reflectance (soil and
383
of images gives then a time series I(t) comprising the values of the selected
spectral index on all selected dates. The I(t)-values give a measure of crop
phenology and can be used for crop identification. To this end images may be
processed to enhance contrast, so that qualitative interpretation and delineation is easier, or numerical classification procedures. may be applied to the
I(x,y,t) images.
C. Calculation and mapping of crop water requirements. Traditionally (F AO,
1977) crop water requirements are estimated by combining climate data with crop
specific coefficients, which accounts for the ratio of the maximum crop specific
evapotranspiration to a reference evapotranspiration dependent on climate only.
More recently (e.g. Allen et aI., 1994) FAO guidelines are evolving towards the
use of the Penman- Monteith (PM) equation which avoids the use of empirical
crop coefficients. The rationale for this approach is that empirical values of crop
coefficients, such as provided by FAO (1977), cannot account for the actual
variability of canopy properties which control consumptive water use. The two
approaches are less different than it may be concluded at a preliminary analysis. It
can be shown, using the PM equation, that the crop coefficients are related
explicitly to the parameters of the PM equation (Stanghellini et aI., 1990).
Accordingly, radiance measurements may contribute in different ways to the
determination and mapping of crop water requirements:
accurate determination of cropped area (see A) reduces significantly the
uncertainty on crop water requirements within an irrigation scheme.
mapping of specific crops (see B) in combination with accurate knowledge of
crop phenology reduces further the uncertainty on the spatial and temporal
variability of crop water requirements.
clusters of crops having similar crop coefficients can be mapped using
advanced numeric classification procedures (D'Urso and Menenti, 1996a, b);
estimates of surface reflectance, leaf area index, surface temperature can be
derived from spectral and directional radiances and applied to compute maps
of crop water requirements.
estimates of net radiation, Rn, and soil heat flux, G, can be obtained from
spectral radiances, with (Rn - G) being an useful upper bound of maximum
evapotranspiration and, therefore, of crop water requirements.
D. Crop stress detection using visible through near infrared radiances.
Spectral reflectance of leaves is very sensitive to water absorption, with the effect
being observable even in the broad bands of Thematic Mapper (Jackson et aI.,
1980). At higher spectral resolutions (e.g. 10 nm), water absorption is observable
at multiple wavelengths (Curran, 1989). Determination ofleafwater content gives
a direct measure of water shortage. Relationships between properties of green
foliage, e.g. leaf water content, and reflectance spectra of canopies change (e.g.
Wessmann, 1992) with canopy architecture and background reflectance (soil and