5 Remote Sensing in Hydrological Modeling
91
junction with GCIP, which is focused within the Mississippi River basin. It was
selected because of the wealth of historical observational data. Nonetheless, there
were only 26 stations within the Red River-Arkansas basin with long-term records
of wind and relative humidity, and even fewer with solar and longwave radiation
observations. Elsewhere globally, the availability of forcing variables over large
areas (especially radiation) is even more limited and will increasingly be a factor
as the emphasis on continental and global modeling increases.
Remote sensing offers a potentially attractive alternative to the use of ground
observations as the forcings for hydrological modeling given:
(1) the recent availability of consistent, long-term remote sensing records, such as
the A VHRR (Agbu, 1993), GOES (Young, 1995) and SSMII (Hollinger et al.,
1992) Pathfinder data sets, as well as other compilations of remote data such
as part of the ISLSCP (Sellers et al., 1995) initiative;
(2) recent advances in remote sensing algorithms for deriving forcing variables
that have been traditionally measured on the ground, such as radiation, humidity, and temperature;
(3) the development of new remote sensing instruments such as rain radar, and the
future NASA Earth Observing System suite of sensors, that may be used to directly or indirectly estimate the required forcing fields, hopefully with increased accuracy.
There is a need to develop a predominantly remote sensing approach to macroscale hydrological modeling. To achieve this, land surface hydrologic models must
be developed that are capable of utilizing remotely-sensed data; and, to develop
and test remote sensing algorithms appropriate for generating data for hydrologic
modeling. Equation (5.1) gave the primitive water balance equation. The corresponding energy balance equation is:
RlI =AE+H +G
(5.2)
where Rn is net surface radiation (solar and longwave), AE is the latent heat, H the
sensible heat and G the ground heat flux. Equations (5.1) and (5.2) are not directly
usable for determining the terrestrial water and energy fluxes and states. To make
LIS
them usable, the flux and state terms ( - , P, E, Q, H and G) must be parameterdt
ized in terms of the state variables which are the soil moisture profile, the surface,
ground and near-surface air temperatures, and near-surface humidity. This parameterization leads to so-called Soil-Vegetation-Atmosphere-Transfer (SVAT) models.
Figure 5.2 shows the data required for such land surface models. The inputs to
the model can be divided into three categories:
(i.) Forcing variables needed to drive the model. These include precipitation
(both liquid and solid), incoming solar (shortwave and near-infrared) radiation from the atmosphere, and downwelling longwave (thermal) radiation
from the atmosphere.
(ii.) Surface meteorology that is required as part of the parameterization of the
evaporation, transpiration and sensible heat variables. The required meteorology includes surface air temperature, surface humidity and surface wind.
91
junction with GCIP, which is focused within the Mississippi River basin. It was
selected because of the wealth of historical observational data. Nonetheless, there
were only 26 stations within the Red River-Arkansas basin with long-term records
of wind and relative humidity, and even fewer with solar and longwave radiation
observations. Elsewhere globally, the availability of forcing variables over large
areas (especially radiation) is even more limited and will increasingly be a factor
as the emphasis on continental and global modeling increases.
Remote sensing offers a potentially attractive alternative to the use of ground
observations as the forcings for hydrological modeling given:
(1) the recent availability of consistent, long-term remote sensing records, such as
the A VHRR (Agbu, 1993), GOES (Young, 1995) and SSMII (Hollinger et al.,
1992) Pathfinder data sets, as well as other compilations of remote data such
as part of the ISLSCP (Sellers et al., 1995) initiative;
(2) recent advances in remote sensing algorithms for deriving forcing variables
that have been traditionally measured on the ground, such as radiation, humidity, and temperature;
(3) the development of new remote sensing instruments such as rain radar, and the
future NASA Earth Observing System suite of sensors, that may be used to directly or indirectly estimate the required forcing fields, hopefully with increased accuracy.
There is a need to develop a predominantly remote sensing approach to macroscale hydrological modeling. To achieve this, land surface hydrologic models must
be developed that are capable of utilizing remotely-sensed data; and, to develop
and test remote sensing algorithms appropriate for generating data for hydrologic
modeling. Equation (5.1) gave the primitive water balance equation. The corresponding energy balance equation is:
RlI =AE+H +G
(5.2)
where Rn is net surface radiation (solar and longwave), AE is the latent heat, H the
sensible heat and G the ground heat flux. Equations (5.1) and (5.2) are not directly
usable for determining the terrestrial water and energy fluxes and states. To make
LIS
them usable, the flux and state terms ( - , P, E, Q, H and G) must be parameterdt
ized in terms of the state variables which are the soil moisture profile, the surface,
ground and near-surface air temperatures, and near-surface humidity. This parameterization leads to so-called Soil-Vegetation-Atmosphere-Transfer (SVAT) models.
Figure 5.2 shows the data required for such land surface models. The inputs to
the model can be divided into three categories:
(i.) Forcing variables needed to drive the model. These include precipitation
(both liquid and solid), incoming solar (shortwave and near-infrared) radiation from the atmosphere, and downwelling longwave (thermal) radiation
from the atmosphere.
(ii.) Surface meteorology that is required as part of the parameterization of the
evaporation, transpiration and sensible heat variables. The required meteorology includes surface air temperature, surface humidity and surface wind.