Improvements in radar rainfall estimation for hydrological modeling | |
| Author | Thanapon Piman |
| Call Number | AIT Diss. no.WM-06-02 |
| Subject(s) | Rain and rainfall--Estimates Hydrologic models |
| Note | A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Engineering in Water Engineering and Management, School of Engineering and Technology |
| Publisher | Asian Institute of Technology |
| Series Statement | Dissertation ; no. WM-06-02 |
| Abstract | Mean areal rainfall estimates are an important input for hydrological and environmental modeling as well as for water resources management. Rain gauges are traditionally used for measuring mean areal rainfall at the ground level. However, inaccurate rainfall estimates based on rain gauges are due to inadequate spatial coverage or configuration and inadequate gauge density, especially in mountainous regions. Estimation of rainfall and runoff in ungauged basins is also a major challenge for hydrologists and water managers due to gauge data being absent, insufficient or of questionable reliability. A promising development in this field is the application of weather radar which measures rainfall with high spatial and temporal resolutions and a large areal coverage, compared to rain gauge networks. However, the radar rainfall measurements also suffer from various types of errors and uncertainties as well, including reflectivity measurement errors, errors in converting reflectivity to rainfall intensity and mean field bias between radar rainfall and rain gauge rainfall, which necessitate correcting or reducing these errors and uncertainties before using the radar measured rainfall data. The objectives of present study are to: develop a method to improve radar-measured rainfall, evaluate the applicability of improved radar rainfall estimates for runoff simulation in a gauged basin, assess the impact of radar and gauge-measured rainfall uncertainties on runoff simulation and model parameters, and estimate rainfall and runoff in an ungauged basin. Two watersheds, Mae Chaem and Mae Klang, located in mountainous regions in the north of Thailand are selected for the study. The radar rainfall data is observed by an S-Band Doppler weather surveillance radar system established at the Om Koi radar site located on the top of a mountain near the study watersheds. The window correlation matching method (WCMM) is developed to reduce collocation and timing errors in matching pairs of radar-measured reflectivity, Ze and gauge-measured rainfall intensity, R for improved estimation of Ze-R relationships and radar rainfall. This method is compared with traditional matching method (TMM), probability matching method (PMM) and window probability matching method (WPMM). The impact of the number of rain gauges on WCMM is investigated to evaluate the sensitivity of a and b parameters in Ze-R relationship and accuracy of radar rainfall estimates. The improved Ze-R relationship obtained from WCMM is then used to estimate mean areal rainfall over the Mae Chaem watershed for hydrologic modeling. The rainfall-runoff simulation is carried out by the quasi-distributed HEC-HMS model based on the hourly data. Furthermore, several uncertainty scenarios based on gauge and radar rainfall estimates are applied to the calibrated model to evaluate their impacts on the accuracy of the simulated runoff. In order to investigate how the calibrated model parameters change with rainfall estimation uncertainties, various uncertainty scenarios are used to re-calibrate the model parameters. The improved radar measured rainfall obtained through WCMM is also examined to estimate point rainfall over 30 rain gauges which are located outside the Mae Chaem watershed. Finally, the calibrated Ze-R relationship and regionalization technique are used to estimate mean areal rainfall and model parameters, respectively in an assumed ungauged basin, the Mae Klang watershed to predict runoff at the outlet of this watershed. In this study, it is found that the calibrated relationship Ze = 18.05R1.45 obtained from WCMM produces the best results for radar rainfall estimates as compared with the other three techniques (TMM, PMM and WCMM). The simulation results indicate that both model calculated runoff volume and peak discharge with the calibrated Ze-R relationship are close to the observed data for both calibration and validation periods. It is found that the uncertainties in gauge- and radar-measured rainfall affect the accuracy of runoff simulation. The runoff estimates with a less number of rain gauges and different geographic locations can be quite different than the actual observed values. The calculated hydrographs using the two Z-R relationships, taken from the literature, based on convective and stratiform rain types provide large underestimation compared to the observed data. The accuracy of runoff simulation is improved when a Z-R relationship based on orographic rain type but developed in another area (Hawiian Islands) is used to estimate radar rainfall in the Mae Chaem watershed compared to those based on the other two relationships. Moreover, it is shown that parameter Sc is largely affected by rainfall estimation uncertainties, followed by parameters Ia and Tc. The parameter K is very sensitive to rainfall estimation uncertainties while, the parameter X varies in a narrow range as compared to the parameter K. The analysis results indicate that the calibrated Ze-R relationship is accurate for a distance of 100-120 km from the center of the Mae Chaem watershed. The calibrated Ze-R relationship and regionalization technique are used to estimate the runoff. The model calculated runoff volume and peak discharge match well with the observed data. |
| Year | 2007 |
| Corresponding Series Added Entry | Asian Institute of Technology. Dissertation ; no. WM-06-02 |
| Type | Dissertation |
| School | School of Engineering and Technology |
| Department | Department of Civil and Infrastucture Engineering (DCIE) |
| Academic Program/FoS | Water Engineering and Management (WM) |
| Chairperson(s) | Babel, Mukand S.; |
| Examination Committee(s) | Oki, Taikan ;Honda, Kiyoshi ;Sutat Weesakul ;Gupta, Ashim Das; |
| Scholarship Donor(s) | RTG Fellowship; |
| Degree | Thesis (Ph.D.) - Asian Institute of Technology, 2006 |