Forecasts Data

Hydrological forecasts are used by many societal sectors, for instance warning services to prepare citizens for floods, farmers to plan agricultural practices (e.g. seeding, harvest, irrigation, fertilization) and reservoir managers to decide on water storage or release. Short-term forecasts are for immediate actions while seasonal forecast affect more strategical decisions and market prices.

The HYPE model provides forecasts daily for 1-10 days ahead and monthly for 1-6 months ahead, also for areas without measurements. The calculations are made on a daily time step and the spatial resolution is given by landscape delineation into catchments, for which HYPE data represents average conditions or the outlets. The forecast is produced by initializing the hydrological model with saved state variables and near-time modelling, before running the HYPE model with meteorological forecasts. The meteorological forecast can be one single deterministic forecast or an ensemble of probabilistic forecasts.

We show both critical areas and provide time-series at specific sites:

Currently, the 1-10 days forecast in HypeWeb is based on one single deterministic meteorological forecast for each geographical domain (from ECMWF). Graphs are showing the river flow modelled with HYPE for the last 3 months and the forecast for the up-coming next 10 days. In most applications, the graph also show return periods for high flows and sometimes, the map is colored based on thresholds in return periods.

Explore 1-10 days forecasts of river flow in graphs by clicking a catchment in the interactive maps below!

For seasonal forecasts, the river flow from HYPE is accompanied with modelled precipitation and temperature to better judge the credibility. For time periods more than ten days ahead, the meteorological models are less reliable. Seasonal forecasts are therefore based on an ensemble of meteorological model results. Currently, we use 50 ensemble members from ECMWF. Instead of absolute values, the maps of 1-6 months forecasts thus show probabilities for river flow in three categories: above normal, near normal and below normal. Regions with low skill can be masked out from the maps, while graphs show the full ensemble range and extreme values for specific sites.

Explore seasonal forecasts in maps by selecting a variable and in graphs by clicking a catchment in the interactive maps below!