The objective of the study was to update and improve the reliability and accuracy of short duration (s 24 h) design rainfall values for South Africa. These were to be based on digitised rainfall data whereas previous studies conducted on a national scale in South Africa were based on data that were manually extracted from autographic charts. With the longer rainfall records currently available compared to the studies conducted in the early 1980s, it was expected that by utilising the longer, digitised rainfall data in conjunction with regional approaches, which have not previously been applied in South Africa, that more reliable short duration design rainfall values could Ix: estimated. A short duration rainfall database was established for South Africa with the majority of the data contributed by the South African Weather Bureau (SAWB). Numerous errors such as negative and zero time steps were identified in the SAWB digitised rainfall data. Automated procedures were developed to identify the probable cause of the errors and appropriate adjustments to the data were made. In cases where the cause of the error could be established, the data were adjusted to introduce randomly either the minimum, average or maximum intensity into the data as a result of the adjustment. The effect of the adjustments was found to have no significant effect on the extracted Annual Maximum Series (AMS). However, the effect of excluding erroneous points or events with erroneous points resulted in significantly different AMS. The low reliability of much of the digitised SAW B rainfall data was evident by numerous and large differences between daily rainfall totals recorded by standard, non-recording raingauges, measured at 08:00 every day, and the total rainfall depth for the equivalent period extracted from the digitised data. Hence alternative techniques of estimating short duration rainfall values were developed, with the focus on regional approaches and techniques that could be derived from daily rainfall totals measured by standard raingauges. Three approaches to estimating design storms from the unreliable short duration rainfall database were developed and evaluated. The first approach used a regional frequency analysis, the second investigated scaling relationships of the moments of the extreme events and the third approach used a stochastic intra-daily model to generate synthetic rainfall series. In the regional frequency analyses, 15 relatively homogeneous rainfall clusters were identified in South Africa and a regional index storm based approach using L-moments was applied. Homogeneous clusters were identified using site characteristics and tested using at-site data. The mean of the AMS was used as the index value and in 13 of the 15 relatively homogeneous clusters the index value for 24 h durations were well estimated as a function of site characteristics only, thus enabling the estimation of 24 h duration design rainfall values at any location in South Africa. In 13 of the 15 clusters the scaling properties of the moments of the AMS were used to successfully estimate design rainfall values for duration < 24h, using the moments of the AMS extracted from the data recorded by standard raingauges and regional relationships based on site characteristics. It was found that L-moments scaled better and over a wider range of durations than ordinary product moments. A methodology was developed for the derivation of the parameters for two Bartlett-Lewis rectangular pulse models using only standard raingauge data, thus enabling the estimation of design values for durations as short as 1 h at sites where only daily rainfall data are available. In view of the low reliability of the majority of short duration rainfall data in South Africa, it is recommended that the regional index value approach be adopted for South Africa, but scaled using values derived from the daily rainfall data. The use of the intra-daily stochastic rainfall models to estimate design rainfall values is recommended as further independent confirmation of the reliability of the design values. / Thesis (M.Sc.)-University of Natal, Pietermaritzburg, 1998.

Tung Wai Lan, Iris. / Thesis (M.Phil.)--Chinese University of Hong Kong, 1990. / Bibliography: leaves 92-101. / ACKNOWLEDGEMENTS / ABSTRACT / LIST OF FIGURES --- p.iii / LIST OF TABLES --- p.iv / CHAPTER / Chapter I --- INTRODUCTION --- p.1 / Chapter 1.1 --- Background --- p.1 / Chapter 1.2 --- Significance of the Research --- p.3 / Chapter 1.3 --- Objectives of the Research --- p.5 / Chapter 1.4 --- Organization of the Research --- p.5 / Chapter II --- LITERATURE REVIEW --- p.7 / Chapter 2.1 --- Introduction --- p.7 / Chapter 2.2 --- Development of Long-Range Forecasting Technique --- p.8 / Chapter 2.3 --- Available Techniques of Long-Range Forecast --- p.9 / Chapter 2.3.1 --- Analogs and persistence --- p.10 / Chapter 2.3.2 --- Statistical modelling --- p.12 / Chapter 2.3.3 --- Atmosphere-ocean interaction --- p.17 / Chapter 2.3.4 --- Cycles and time series --- p.18 / Chapter 2.3.5 --- Numerical modelling --- p.19 / Chapter 2.4 --- Rainfall Prediction in Hong Kong --- p.21 / Chapter III --- RAINFALL OF HONG KONG --- p.24 / Chapter 3.1 --- Climatic Feature --- p.24 / Chapter 3.2 --- The Causes of Hong Kong Rainfall --- p.26 / Chapter 3.2.1 --- Tropical cyclone --- p.26 / Chapter 3.2.2 --- Trough or front --- p.28 / Chapter IV --- METHODOLOGY --- p.31 / Chapter 4.1 --- Introduction --- p.31 / Chapter 4.2 --- Empirical Orthogonal Function (EOF) Analysis --- p.32 / Chapter 4.2.1 --- What's EOF --- p.32 / Chapter 4.2.2 --- Why use EOF --- p.34 / Chapter 4.3 --- Discriminant Analysis --- p.36 / Chapter 4.4 --- Data Base --- p.37 / Chapter 4.5 --- Computation Procedures --- p.40 / Chapter 4.6 --- Analysis of Forecast Capability --- p.44 / Chapter V --- THE RESULT AND ANALYSIS OF PREDICTION MODEL --- p.48 / Chapter 5.1 --- The result of EOF analysis --- p.48 / Chapter 5.1.1 --- Extraction of eigenvectors and eigenvalues --- p.48 / Chapter 5.1.2 --- Spatial and Temporal variation of eigenvector pattern --- p.52 / Chapter 5.2 --- Accuracy of the prediction model --- p.53 / Chapter 5.2.1 --- Introduction --- p.53 / Chapter 5.2.2 --- The forecast accuracy from each month --- p.54 / Chapter 5.2.2.1 --- The forecast accuracy made by October --- p.54 / Chapter 5.2.2.2 --- The forecast accuracy made by November --- p.56 / Chapter 5.2.2.3 --- The forecast accuracy made by December --- p.58 / Chapter 5.2.2.4 --- The forecast accuracy made by January --- p.58 / Chapter 5.2.2.5 --- The forecast accuracy made by February --- p.61 / Chapter 5.2.2.6 --- The forecast accuracy made by March --- p.61 / Chapter 5.2.2.7 --- The forecast accuracy made by April --- p.64 / Chapter 5.2.3 --- Optimal length of dependent data --- p.64 / Chapter 5.2.4 --- Analysis the prediction results --- p.67 / Chapter 5.2.5 --- Comparison between the method used in this study with those methods adopted by ROHK --- p.69 / Chapter 5.2.5.1 --- Introduction --- p.69 / Chapter 5.2.5.2 --- Comparison of the forecast accuracy between two studies --- p.70 / Chapter VI --- CONCLUSION --- p.73 / Chapter 6.1 --- Summary of Findings --- p.73 / Chapter 6.2 --- Limitations of the Research --- p.75 / Chapter 6.3 --- Prospects of the Research --- p.76 / APPENDICES --- p.78 / LIST OF CITED REFERENCES --- p.92 / LIST OF READING MATERIALS --- p.97

Distributed Rainfall-Runoff models are gaining widespread acceptance; yet, a fundamental issue that must be addressed by all users of these models is definition of an acceptable level of watershed discretization (geometric model complexity). The level of geometric model complexity is a function of basin and climatic scales as well as the availability of input and verification data. Equilibrium discharge storage is employed to develop a quantitative methodology to define a level of geometric model complexity commensurate with a specified level of model performance. Equilibrium storage ratios are used to define the transition from overland to channel -dominated flow response. The methodology is tested on four subcatchments in the USDA -ARS Walnut Gulch Experimental Watershed in Southeastern Arizona. The catchments cover a range of basins scales of over three orders of magnitude. This enabled a unique assessment of watershed response behavior as a function of basin scale. High quality, distributed, rainfall -runoff data was used to verify the model (KINEROSR). Excellent calibration and verification results provided confidence in subsequent model interpretations regarding watershed response behavior. An average elementary channel support area of roughly 15% of the total basin area is shown to provide a watershed discretization level that maintains model performance for basins ranging in size from 1.5 to 631 hectares. Detailed examination of infiltration, including the role and impacts of incorporating small scale infiltration variability in a distribution sense, into KINEROSR, over a range of soils and climatic scales was also addressed. The impacts of infiltration and channel losses on runoff response increase with increasing watershed scale as the relative influence of storms is diminished in a semiarid environment such as Walnut Gulch. In this semiarid environment, characterized by ephemeral streams, watershed runoff response does not become more linear with increasing watershed scale but appears to become more nonlinear.

MENVSC (Geography) / Department of Geography and Geo-Information Sciences / Rainfall is undoubtedly the most significant factor for life’s continuity. South Africa is prone to future climate uncertainties due to global climate change. The aim of this study is to investigate rainfall variability and change in South Africa on a present day (1976-2005), near-future (2006-2035) and far-future (2036-2065) climate. For the study, 3 RCMs (REMO2009, RCA4 and CCLM4-8-17), forming part of CORDEX-Africa project were nested within 5 different CIMP5_GCMs of low resolution. GPCC precipitation, NOAA GHCN_CAMS Land Temperature and other NCEP reanalysis products were useful in validating models in simulations of present-day climate. RCP4.5 and RCP8.5 emission scenarios from IPCC-AR5 were used for future climate projections. On the validation, each regional climate model displayed different signature on simulations, rainfall in particular because this is a variable that is affected most by sub-grid process. Simulations nested within MIROC5 simulated more precipitation than simulations forced with other GCMs, due to more large-scale moisture convergence into the nested domain. There were differences in projections of RCM nested within the same GCM, as well as with the same RCM nested within different GCMs, on the future. Models nested within MPI project wetter conditions over the eastern parts of Limpopo, while the other two projected drier conditions in the same area. REMO2009 forced on MPI uniquely projected drying of Western Cape throughout the seasons on both RCPs and futures. Simulations conducted with the RCP8.5 scenario forcing are generally found to be associated with either a larger increase in temperature, or an increase in area associated with higher temperature increases. CCLM4-8-17 forced on HadGEM2 projected below average temperatures over the northwest parts of the country under the RCP8.5 scenarios. MPI driving model projected a general reduction of evaporation values, with lowest over northeast, northwest parts and south coastal parts of South Africa, in contrary to adjacent oceans. In this study, we have sought to identify the sources of uncertainties amongst model simulations between either the RCMs or the driving GCMs. / NRF

Text in Afrikaans, abstract in Afrikaans and English / As deel van die weermodifikasie-eksperiment in Bethlehem, Suid-Afiika, is 'n reenmeternetwerk geinstalleer, en word die neerslagwaardes R; wat by 43 reenmeterstasies waargeneem is, vergelyk met die waargenome radar reflektiwiteit Z;. Alhoewel radar ruimtelike en tydskontinue metings van reflektiwiteit bied wat onmiddellik by een sentrale punt beskikbaar is, is die akkuraatheid van radar om reenval te meet onseker as gevolg van verskeie potensiele foute in die omskakeling van reflektiwiteit na reenval. Dit word aanvaar dat reenmeters akkurate puntwaarnemings van reenval gee en daar bestaan eenstemmigheid dat die kombinasie van die twee metodes beter is as enigeen van die metodes afsonderlik. In hierdie studie ondersoek ek die toepassing van die veralgemeende lineere model as 'n beramingstegniek. Vorige studies gebruik die log-log transformasie, d. w.s. logZ = logA + b(logR) van die Z = ARb verwantskap om die koeffisiente A en b met behulp van kleinste-kwadrate-regressie te bepaal. Die implisiete aanname hiermee is dat die foute ongekorreleerd is. Met die inverse verwantskap R = czd d.w.s. logR = logC + d(logZ) neem ek aan dat die waarnemings nie onafhanklik is nie sodat die regressiekoeffisiente bereken word met behulp van die metode van die veralgemeende lineere model. Om die ruimtelike afhanklikheid van die reenmeterwaarnemings te modelleer, word eksperimentele variogramme uit die data bereken en gepas met teoretiese variogramme wat gebruik word om die variansie-kovariansiematriks te vu!. "Gemiddeld" vaar hierdie metode beter as gewone regressie vir analises wat reenmeters wat verder as 45km vanaf die radarstel is, insluit. Residu-stipping wys dat die afstand van die meter vanaf die radarstel as 'n afsonderlike onafhanklike veranderlike in die regressievergelyking ingesluit behoort te word, d.w.s. die beraming verbeter met logR = 3-0 + a,(logZ) + a2(afstand). Hierdie meervoudige regressiemodel stem ooreen met die teoretiese model van Smith en Krajewski omdat e -- afstand as 'n praktiese manifestasie van die foutproses [e.,, (ij)] beskou kan word. Omdat E(ez) = eE<ZJ e'"a' as Z 'n lognormaalverdeling het, kan die sydigheid wat ontstaan wanneer antilogaritmes geneem word, reggestel word deur die beraamde reenval met e112 "' te vermenigvuldig. Die studie !ewer 'n bydrae met die afleiding van 'n beramingstegniek wat die beraming van neerslag uit radar betekenisvol verbeter. / In a study of a rain-gauge network that was installed for a weather modification experiment in Bethlehem, South Africa, precipitation values R; observed at 43 gauging stations are compared to the observed radar reflectivity Z;. Although radar provides spatial and temporal measurements of reflectivity that are immediately available at one location, the accuracy of radar estimation of rainfall is uncertain due to various potential errors in the conversion from reflectivity to rainfall. Rain-gauges are assumed to give accurate point measurements of rainfall and there is general agreement that the combination of systems is better than either system alone. In this study I explore the application of the general linear model as an estimation technique. Previous studies have used the log-log transform, i.e. logZ = logA + b(logR) of the Z = ARb relation, and applied least-squares regression analysis to determine the coefficients A and b. This implicitly assumes that the disturbances are uncorrelated. Working with the inverse relation R = czd i.e. logR = logC + d(logZ) and assuming that the observations are not independent we compute the regression coefficients using generalised least squares. To model the spatial dependence of the rain-gauge observations we compute experimental variograms from the data and fit them with theoretical variograms which are then used to fill the variance-covariance matrix. "On average" this method performs better than ordinary regression for the analyses that included rain-gauges further than 45km from the radar set. Residual plotting revealed that distance of the rain-gauge from the radar set should be included as a separate independent variable in the regression equation, i.e. logR = ao + a1(logZ) + a1(distance) improved the estimation of rainfall as it performs better than ordinary regression. This multiple regression model agrees with the theoretical model of Smith and Krajewski in the sense that e "'distance is a practical manifestation of the error process [ e,, (ij)]. Showing that E( ez) = el!.(!.) e 112 "' if Z has a lognormal distribution, the bias when taking antilogs can be removed by multiplying estimated rainfall by e1 ' 2a'. The contribution of this study is the derivation of an estimation technique which significantly improves the estimation of rainfall from radar / Mathematical Sciences / D. Phil. (Statistics)

Infrared (IR) imagery collected by geostationary satellites provides useful information about the dirunal evolution of cloud systems. These IR images can be analyzed to indicate the location of clouds as well as the pattern of cloud top temperatures (Tbs). During the past several decades, a number of different approaches for estimation of rainfall rate (RR) from Tb have been explored and concluded that the Tb-RR relationship is (1) highly nonlinear, and (2) seasonally and regionally dependent. Therefore, to properly model the relationship, the model must be able to: (1) detect and identify a non-linear mapping of the Tb-RR relationship; (2) Incorporate information about various cloud properties extracted from IR image; (3) Use feedback obtained from RR observations to adaptively adjust to seasonal and regional variations; and (4) Effectively and efficiently process large amounts of satellite image data in real -time. In this study, a kind of artificial neural network (ANN), called Modified Counter Propagation Network (MCPN), that incorporates these features, has been developed. The model was calibrated using the data around the Japanese Islands provided by the Global Precipitation Climatology Project (GPCP) First Algorithm Intercomparison Project (AIP-I). Validation results over the Japanese Islands and Florida peninsula show that by providing limited ground-truth observation, the MCPN model is effective in monthly and hourly rainfall estimation. Comparison of results from MCPN model and GOES Precipitation Index (GPI) approach is also provided in the study.

published_or_final_version / Civil Engineering / Doctoral / Doctor of Philosophy

Measurements of temperature and salinity in the upper 5 m of the ocean along the equator showed cool fresh anomalies due to rain showers. The measurements were made between 140 W and 110 W during April 1987, an El Nino year. The eastern equatorial Pacific was characterized by weak winds (3 m/s average), high rainfall (1.6 cm/day), and warm surface temperatures (28.4 C). Measurements of temperature were made from a catamaran float at 0.5 and 1 m depth and at 5 m depth from the ship. Salinity was measured at a depth of 1 m from the float and 5 m from the ship. The float was towed off of the port side of the ship outside of the bow wake. Near-surface low temperature and low salinity anomalies due to cool rainfall were encountered. These anomalies were on average cool and fresh by 0.02 C and 0.2 PSTJ with maximum values of 0.5 C and 1.6 PSU. The horizontal extent of the anomalies ranged from less than 10 to more than 100 km. Rainfall depths estimated from salt conservation agreed roughly with shipboard rain-gauge measurements. The characteristic lifetime of the anomalies, estimated from the ratio of the average rain depth to average rain rate, was about 10 hrs. Rainfall temperatures were computed from the T-S mixing curves for three large, newly-formed anomalies. The average rainfall temperature was 21 C. Ocean buoyancy fluxes estimated for intense rain showers were an order of magnitude larger than the fluxes in the absence of rain. / Graduation date: 1992 / Best scan available for p.15-16, 35. Original is a black and white photocopy.

This study is concerned with design forecasting of storm hydrographs with emphasis on runoff volume and peak discharge. The objective of the study was to develop, calibrate and test a method for forecasting storm runoff from small semi-arid watersheds using an available prediction model. In order to turn the selected prediction model into a forecasting model an objective procedure in terms of an API-type model was developed for evaluating the soil moisture deficit in the upper soil layer at the beginning of each storm. Distinction was made between the physically-based parameters and the other fitting parameters. The rainfall excess calculation was computed by solving the Green and Ampt equation for unsteady rainfall conditions using the physically-based parameters. For the physically-based parameters a geographic information system was developed in order to account for the variability in time and space of the input data and the watershed characteristics and to coregister parameters on a common basis. The fitting parameters were used to calibrate the model on one subwatershed in the Walnut Gulch Experimental Watershed while the physically-based parameters remained constant. Two objective functions were selected for the optimization procedure. These functions expressed the goodness of fit between the calculated hydrograph volume and peak discharge and the observed volume and peak discharge. Linear relationships between the effective matric potential parameter and the two objective functions obtained from the sensitivity analyses made it possible to develop a bilinear interpolation algorithm to minimize, simultaneously, the difference between the calculated and observed volume and peak discharge. The prediction mode of the model was tested both on different storm events on the same subwatershed and on another subwatershed with satisfactory results. In the prediction mode the effective matric potential parameter was allowed to vary from storm to storm, however, in the forecasting mode these values were obtained from the API model. Relatively poor results were obtained in testing the forecasting mode on another subwatershed. These errors were able to be corrected by changing the channel losses fitting parameters.

The U.S. Geological Survey and the City of Tucson, Arizona, have been collecting rainfall and runoff data on several watersheds in the Tucson area for several years. Among the purposes of this project is to use the data to test rainfall-runoff models in an effort to find one to successfully simulate flood flows in Tucson. One such model, the Distributed Routing Rainfall-Runoff Model (DR3M), was tested using data collected on Rob Wash in Tucson. It was found DR3M performs about as well as it does in other parts of the United States, although it tends to underestimate flood flows for large storms and overestimate flows for smaller storms. Unique features with regard to the hydrology of urban Tucson require special attention when using DR3M; these features are associated with the nature of dry washes and summer rainfall in Tucson. Experience indicates DR3M is not truly a deterministic model.