Figures index

From

Moving Horizon vs. Unscented Kalman Filter for State Estimation in Streamflow Prediction

Divas Karimanzira, Thomas Rauschenbach

World Journal of Environmental Engineering. 2022, 7(1), 1-10 doi:10.12691/wjee-7-1-1
  • Figure 1. Topology of the Truse catchment area
  • Figure 2. Structure of the conceptual model according to Lorent/Gevers
  • Figure 3. Calibration results for the Lorent/Gevers conceptual model of the Truse catchment.
  • Figure 4. Validation results for the Lorent/Gevers conceptual model of the Truse catchment
  • Figure 5. Effective Rainfall PE(k) of the Lorent/Gevers model in dependence of the free storage (Smax-S(k)) and the total rainfall PB(k) (assuming ETP(k)=0)
  • Figure 6. MHE: Estimated discharge of the catchment
  • Figure 7. MHE: Estimated soil storage content of the catchment
  • Figure 8. MHE: Estimated state of the slow first order lag element of surface flow
  • Figure 9. MHE: Estimated state of the fast first order lag element of the base flow
  • Figure 10. Composition of the catchment outflow from the different routing elements
  • Figure 11. Catchment outflow from the two estimators
  • Figure 12. Estimated soil moisture content
  • Figure 13. Estimated base flow component
  • Figure 14. Estimated lower base flow
  • Figure 15. estimated upper surface flow
  • Figure 16. Estimated slow surface flow
  • Figure 17. Direct estimation of the catchment outflow based on the measured data. a) simulated signal is back propagated, b) measured signal is back propagated
  • Figure 18. Results of adjusting precipitation input. a) Estimated parameters (θ1 and θ2) for adjustment of precipitation input, b) Estimated catchment outflow resulting from the moving horizon state estimator based on the real flow measurements