| Description |
1 online resource (144 p.) |
| Contents |
Cover -- Half Title -- Title Page -- Copyright Page -- Dedication Page -- Abstract -- Contents -- 1 Introduction -- 1.1 Data assimilation -- 1.2 Review of different data assimilation methods -- 1.2.1 Techniques from estimation theory -- 1.2.2 Techniques from control theory -- 1.3 Thesis outline -- 2 The Deterministic Hydrodynamic model -- 2.1 Introduction -- 2.2 The 2D shallow water equations model -- 2.2.1 Numerical Solution -- 2.3 Initial definition of the state vector in MIKE 21 -- 2.4 The 3D shallow water equations model -- 2.4.1 Numerical solution -- 3 Kalman filter techniques |
|
3.1 Linear filtering -- 3.1.1 Sequential updating algorithm -- 3.1.2 State vector augmentation: time coloured noise -- 3.2 Extended Kalman filter -- 3.3 The Reduced Rank SQuare RooT filter -- 3.3.1 The RRSQRT algorithm -- 3.4 The Ensemble Kalman filter -- 3.4.1 The Ensemble Kalman filter scheme -- 3.5 Integration of a DA technique in MIKE21 -- 3.5.1 Definition of the state vector -- 3.5.2 Measurements and measurement noise definition -- 3.5.3 The definition of the system noise -- 3.5.4 Scaling problem for a multivariate state vector -- 3.5.5 Memory reduction factor |
|
4 Study and comparison: the RRSQRT filter and the EnKF -- 4.1 Definition of the true model and the observations -- 4.2 TEST 1: Errors in the open boundary conditions -- 4.2.1 Perfectly known statistics -- 4.2.2 Sensitivity and robustness -- 4.3 TEST 2: Error in the meteorological forcing terms -- 4.3.1 Perfectly known statistics -- 4.3.2 Sensitivity and robustness -- 4.4 Some remarks about the mass conservation problem -- 4.5 Conclusions -- 5 Application on a 2D hydrodynamic model. Test case -- 5.1 The North Sea model test case -- 5.1.1 The model set-up -- 5.1.2 Deterministic model performance |
|
5.1.3 Filter set-up -- 5.1.4 Hindcast results -- 5.1.5 Forecast results -- 5.2 Conclusions -- 6 Application on a 2D Nested hydrodynamic model. Test case -- 6.1 Special features of the Nested model -- 6.2 Implementation of the RRSQRT filter -- 6.3 A test case: a regional model of the North sea and Baltic sea and a nested area in the Danish Inner waters -- 6.4 Conclusions -- 7 Initial tests on a 3D hydrostatic model -- 7.1 Implementation in MIKE 3 HDHS -- 7.2 Twin test -- 7.2.1 Test 1: Systematic error at the open boundary |
|
7.2.2 Test 2: Systematic error at the meteorological forcing terms (No wind) -- 7.3 Conclusions -- 8 Summary and conclusions -- List of Figures -- List of Tables -- Bibliography -- Samenvatting (Summary in Dutch) -- Acknowledgements -- Curriculum Vitae |
| Notes |
Description based upon print version of record |
| Form |
Electronic book
|
| ISBN |
9781000658170 |
|
1000658171 |
|
