the spatial resolution of satellite precipitation is around 5–25 km from the Tropical
Rainfall Measuring Mission (TRMM) at near global coverage. The combined
regional and global precipitation products can be synergistically developed into
5-min products from radar networks and every 3-h products from satellites, which
enables users to accumulate to longer time scales such as 6 h, 12 h, daily, monthly,
annually, and up to decade scales for a variety of researches and applications.
13.3 PRECIPITATION UNCERTAINTY QUANTIFICATION AND
ERROR PROPAGATION INTO HYDROLOGICAL PREDICTION
As a key forcing variable to the land surface hydrological process, it is vital to better
understand precipitation uncertainty quantification and error propagation into hydrological prediction as a function of scales. This section is mainly based on Hong et al.
(2006), which developed an end-to-end uncertainty propagation analysis framework
to first quantify the scale-dependent satellite rainfall uncertainty and further assess its
error propagation into hydrological modeling and stream flow predictions.
13.3.1 Uncertainty Quantification Framework
Uncertainty quantification is defined as a function of the space–time–intensity of
precipitation as shown in Figure 13.1. To quantify the influence of estimation
uncertainty on hydrological responses by ensemble simulation would further place
an appropriate degree of confidence from data users’ perspectives. The advantage of
this error formula, for example, is that the various users of the precipitation products
can approximate the error by directly implementing the function at their specified
space–time scales and rain intensities and quantifying the error propagation into a
conceptual rainfall–runoff hydrological model through Monte Carlo simulation. By
flowing the framework as shown in Figure 13.1, an ensemble of satellite precipitation
FIGURE 13.1 A conceptual “end-to-end” error analysis framework for precipitation data
producers and users.
PRECIPITATION UNCERTAINTY QUANTIFICATION AND ERROR PROPAGATION
255
Rainfall Measuring Mission (TRMM) at near global coverage. The combined
regional and global precipitation products can be synergistically developed into
5-min products from radar networks and every 3-h products from satellites, which
enables users to accumulate to longer time scales such as 6 h, 12 h, daily, monthly,
annually, and up to decade scales for a variety of researches and applications.
13.3 PRECIPITATION UNCERTAINTY QUANTIFICATION AND
ERROR PROPAGATION INTO HYDROLOGICAL PREDICTION
As a key forcing variable to the land surface hydrological process, it is vital to better
understand precipitation uncertainty quantification and error propagation into hydrological prediction as a function of scales. This section is mainly based on Hong et al.
(2006), which developed an end-to-end uncertainty propagation analysis framework
to first quantify the scale-dependent satellite rainfall uncertainty and further assess its
error propagation into hydrological modeling and stream flow predictions.
13.3.1 Uncertainty Quantification Framework
Uncertainty quantification is defined as a function of the space–time–intensity of
precipitation as shown in Figure 13.1. To quantify the influence of estimation
uncertainty on hydrological responses by ensemble simulation would further place
an appropriate degree of confidence from data users’ perspectives. The advantage of
this error formula, for example, is that the various users of the precipitation products
can approximate the error by directly implementing the function at their specified
space–time scales and rain intensities and quantifying the error propagation into a
conceptual rainfall–runoff hydrological model through Monte Carlo simulation. By
flowing the framework as shown in Figure 13.1, an ensemble of satellite precipitation
FIGURE 13.1 A conceptual “end-to-end” error analysis framework for precipitation data
producers and users.
PRECIPITATION UNCERTAINTY QUANTIFICATION AND ERROR PROPAGATION
255