Improved spring flood forecasts

In today's system for spring flood forecasts, the HBV model is run using observed data (precipitation, temperature) to a starting state before the spring flood begins. This makes the model state optimally reflect the actual hydrological state, e.g. with respect to snow cover and soil moisture. Then the weather evolutions during the spring flood period from historical years are used to calculate the spring flood.

This project contains two main activities

  • to optimise today's procedure, for example by using only a subset of the historical weather ensemble or by using ordinary (short-term) meteorological forecasts
  • to investigate whether meteorological long-term forecasts (climate forecasts) can improve the result and, if so, develop a new alternative procedure

Period: 2009-2011

Financing: Elforsk/HUVA

Responsible at SMHI: Jonas Olsson

Partner: Lunds Universitet