faty
Get Complete Project Material File(s) Now! »
Spatial distribution of Biases and Root Mean Square Error
The biases and root mean square error calculated using Eqns 3 and 4 are used to assess how well the RCMs can simulate long term (1971-2005) TN and TX over different stations, while the normalized bias (NBias) and root mean square error (NRMSE) calculated using Eqs 6 and 7 are used to assess how well the RCMs can simulate long term (1971-2005) rainfall over different stations. Figure 2 presents the biases in TN at different stations. From this figure RACMO22T driven by ICHEC underestimates TN by -2°C to zero over central, parts of the northeastern highlands, northern, northern coast and western regions. It simulates TN with biases in the range of -4 to -2°C over Same, Morogoro, Tabora, Iringa, and Songea. This model simulates TN with coldest biases in the range of -5 to -4°C over Moshi and Ulanga (Ilonga) regions. Warm biases of zero to 2°C are simulated over coastal regions and over southwestern highlands. HIRHAM5 driven by ICHEC overestimates TN by zero to 2°C over central, parts of northeastern highlands, Morogoro, Iringa, Tabora and over northern coast regions and by 2 to 4°C, over western, southern, southwestern highlands, coastal and northern regions. It simulates TN with cold biases in the range of -2°C to zero over Moshi and Ulanga (Ilonga) regions.
Interannual variability of TN, TX and rainfall
Outputs from CORDEX RCMs driven by ERA-Interim data are used to assess the ability of the RCMs to simulate inter-annual variations of TN, TX and rainfall. These data are available for the period of 1989-2008. The performance of the models in simulating interannual variability in TN and TX is presented in Figure 10 (a), (b). This figure shows that all RCMs captured the pattern of interannual variability in TN and TX well. The magnitude of TN is overestimated by all RCMs. All RCMs including the ensemble average underestimate the magnitude of TX. The performance of CORDEX RCMs in simulating interannual variability of rainfall is analyzed by comparing simulated and observed interannual variations in rainfall in bimodal and unimodal regions (see Figure 11 (a), (b)). All RCMs except HIRHAM5 reproduce the interannual variability of rainfall in unimodal regions. Furthermore, all RCMs except RCA4 reproduce the interannual variability of rainfall in bimodal regions. It can be concluded that CORDEX RCMs driven by ERA-interim data fairly capture the interannual variability of rainfall in unimodal and bimodal regions.
Chapter 1: Introduction
1.1 Background of the study
1.2 Problem
1.3 Research aim and objectives
1.4 Research Hypothesis and Questions
1.5 Motivation and significance of the study
1.6 Conceptual frame work of the study
1.7 The outline of the thesis
References
Chapter 2: Evaluation of the performance of CORDEX Regional Climate Models in simulating present climate conditions of Tanzania
Abstract
2.1 Introduction
2.2 Data and Methodology
2.3 Results
2.4 Conclusions
Acknowledgments
References
Chapter 3: Assessment of the impacts of climate change on maize production in the Wami Ruvu basin of Tanzania
3.1 Introduction
3.2 Data and Methodology
3.3 Results
3.4 Discussion
3.5 Conclusions and recommendations
Chapter 4: Evaluation of the use of moist potential vorticity and moist potential vorticity vector in describing rainfall events in Tanzania
4.1 Introduction
4.2 Background information on potential vorticity
4.3 Data and analysis
4.4 Results
4.5 Summary and recommendations
Acknowledgement
Reference
Chapter 5: Moist potential vorticity vector for diagnosis of heavy rainfall events in Tanzania Abstract
5.1 Introduction
5.2 Data and Methodology
5.3 Results and discussion
5.4 Conclusion and recommendations
Acknowledgments
Reference
Chapter 6: Summary and recommendations
Reference
