On recognizing uncertainty in hydrological forecasting: mapping
Pappenberger and Beven’s code of practice, principles, and the way
forward
Abstract
In the last decade, recognizing and reducing uncertainties in
hydrological forecasting has shown renewal interest. However, from a
modeler’s perspective, a unified code of practice is always needed to
handle the various facets of uncertainty in hydrological forecasting.
Pappenberger and Beven, (2006) suggested nine codes of practice for
handling uncertainties in hydrological modelling. In this paper, we have
revisited those principles and added new insights to yield seven key
principles for accounting and reducing uncertainties in catchment
related hydrological forecasting tasks: (1) objectives define the need
for uncertainty, (2) exploring the Catchment Puzzle, (3) selection of
models is key, (4) choices of the method for quantifying uncertainties
and calibration (5) finding the sources of uncertainties (6)
advancements are a critical choice (7) prioritizing End User Needs for
Reliable Forecasting Services. We derive these principles as a summary
of understanding how modelers across the world have approached
uncertainty handling from the analysis of recent literature on reducing
uncertainties in hydrological forecasting. The triangulated
interdependence and uncertainty contributions between the hydrological
processes, epistemic uncertainties, and model development inevitably
impact the forecast. Yet, the mapping of these principles provided in
this study can assist the modelers in developing an improved framework
for hydrological forecasting. Further, this work calls for discussions
among the hydrological science community to establish these principles.