Introduction

Question 1

Forecast (real-time forecast):

Answer 1

Estimation of a hydrometeorological variable for a specific time or a specific time period

Question 2

Prognosis:

Answer 2

Estimation of a variabel for the future without relation to a specific time. This estimation is often associated with a probability.

Question 3

Lead time

Answer 3

Difference between time when forecast is issued and future time for which variable is estimated. The forecast horizon is the maximum lead time (see Fig. 8.1).

Question 4

Applications:

Answer 4

• Weather forecast (temperatur, precipitation, radiation, wind, etc.) for leisure, farming, water resources management, traffic, flight navigation, aviagion, etc. → meteorology
• Streamflow forecast (floods, droughts, inflow for dams) for water ressources management, navigation, hydro power, ecology, etc. → hydrology
• Water quality forecast (temperatur, O2-concentration, toxics for water ressources management, cooling water, ecology, disaster warning, etc. → water ressources management

Question 5

Classification acc. to lead time:

Answer 5

1. Shortest time forecast (Nowcasting):
   - Lead time from minutes to a few hours
   - Application for heavy storms, flash floods
2. Short time forecast:
   - Lead time from hours to a few days
   - Application e.g. for flood forecasting
3. Medium term forecast:
   - Lead time from days to 2 weeks
   - E.g. meteorological circulation pattern
4. Long term forecast:
   - Lead time from weeks to months
   - Application e.g. for low flow forecast

Question 6

Ensemble-prediction:
(EPS-Ensemble-Prediction-System):

ensemble = a group of items viewed as a whole rather than individually.

Answer 6

 Because of non-linearity of the systems and uncomplete knowledge an exact forecast is not possible

 Different alternative forecasts can be used to assess the uncertainty

 Ensembles are a set of alternative weather forecasts, valid for the same place and the same time

 Uncertainty can be estimated by considering the frequency of certain states (e.g. 80% probability of precipitation)

 Hydrological models can be driven with meteorological ensembles → ensemble flood prediction, gives e.g. uncertainty bands for discharge hydrographs or probabilities of threshold exceedings

Question 7

Ensembles can be built by:

ensemble = a group of items viewed as a whole rather than individually.

Answer 7

a) Running several alternative models
b) Running the same model with different initial/ boundary conditions or model parameters
c) Combination of forecasts with different