Short-term stream flow forecasting and downstream gap infilling using machine learning techniques
| dc.contributor.advisor | Smit, G. J. F. | en_ZA |
| dc.contributor.advisor | Brink, Willie | en_ZA |
| dc.contributor.advisor | Wilms, Josefine M. | en_ZA |
| dc.contributor.author | Steyn, Melise | en_ZA |
| dc.contributor.other | Stellenbosch University. Faculty of Science. Dept. of Mathematical Sciences (Applied Mathematics) | en_ZA |
| dc.date.accessioned | 2018-02-15T14:36:17Z | |
| dc.date.accessioned | 2018-04-09T06:54:52Z | |
| dc.date.available | 2018-02-15T14:36:17Z | |
| dc.date.available | 2018-04-09T06:54:52Z | |
| dc.date.issued | 2018-03 | |
| dc.description | Thesis (MSc)--Stellenbosch University, 2018. | en_ZA |
| dc.description.abstract | ENGLISH ABSTRACT : Stream flow is an important component in the hydrological cycle and plays a vital role in many hydrological applications. Accurate stream flow forecasts may be used for the study of various hydro-environmental aspects and may assist in reducing the consequences of floods. The utility of time series records for stream flow analyses is often dependent on continuous, uninterrupted observations. However, interruptions are often unavoidable and may negatively impact the sustainable management of water resources. This study proposes the application of machine learning techniques to address these hydrological challenges. The first part of this study focuses on single station short-term stream flow forecasting for river basins where historical time series data are available. Two machine learning techniques were investigated, namely support vector regression and multilayer perceptrons. Each model was trained on historical stream flow and precipitation data to forecast stream flow with a lead time of up to seven days. The Shoalhaven, Herbert and Adelaide rivers in Australia were considered for experimentation. The predictive performance of each model was determined by the Pearson correlation coefficient, the root mean squared error and the Nash-Sutcliffe efficiency, and the predictive capabilities of the models were compared to that of a physically based stream flow forecasting model currently supplied by the Australian Bureau of Meteorology. Based on the results, it was concluded that the machine learning models have the ability to overcome certain challenges faced by physically based models and the potential to be useful stream flow forecasting tools in river basin modelling. The second part of this study investigates the ability of support vector regression and multilayer perceptron models to infill incomplete stream flow records. The infilling techniques relied upon data from donor stations and rain gauges within close proximity to the station considered for infilling. A case study was conducted on a channel in the Goulburn basin in Australia. The results showed the promising role of machine learning applications for the infilling of gaps in stream flow records and indicated that data from donor stations contribute more to the success of these models compared to precipitation data. | en_ZA |
| dc.description.abstract | AFRIKAANSE OPSOMMING : Stroomvloei is ’n belangrike komponent in die hidrologiese siklus en speel ’n prominente rol in verskeie hidrologiese toepassings. Akkurate stroomvloeivoorspellings kan vir die bestudering van verskeie hidrologiese omgewingsaspekte gebruik word en kan help om die nagevolge van vloede te verminder. Die gebruik van tydreeksdata vir stroomvloei-analise is dikwels afhanklik van ononderbroke waarnemings. Onderbrekings is egter dikwels onvermydelik en kan ’n negatiewe impak op die volhoubare bestuur van waterhulpbronne hê. In hierdie studie is die toepassing van masjienleertegnieke met die doel om hierdie hidrologiese uitdagings aan te spreek, bestudeer. In die eerste gedeelte van hierdie studie is daar op korttermyn stroomvloeivoorspellings by meetstasies wat oor beskikbare historiese tydreeksdata beskik, gefokus. Twee masjienleertegnieke is ondersoek, naamlik steunvektor-regressie en multi-laag perseptron modelle. Elke model is op historiese stroomvloei- en reënvaldata afgerig om stroomvloei tot en met sewe dae vooruit te voorspel. Eksperimente is op die Shoalhaven, Herbert en Adelaide riviere in Australië uitgevoer. Die voorspellingsvermoëns van elke model is deur die Pearsonkorrelasiekoëffisiënt, die wortel-gemiddelde-kwadraat fout en die Nash-Sutcliffedoeltreffendheid bepaal, en is met dié van ’n fisiese stroomvloeivoorspellingsmodel wat tans deur die Australiese Buro vir Meteorologie verskaf word, vergelyk. Op grond van die resultate is daar tot die gevolgtrekking gekom dat die masjienleermodelle oor die vermoëns beskik om sekere uitdagings waarmee fisiese modelle gekonfronteer word, te oorkom, en dat hulle ’n waardevolle bydrae tot die modellering van riverkomme kan lewer. In die tweede gedeelte van hierdie studie is steunvektor-regressie en multilaag perseptron modelle se vermoëns om onvolledige stroomvloeistate te vul, ondersoek. Die invultegnieke was afhanklik van data vanaf ander nabygeleë meetstasies en reënmeters. ’n Gevallestudie is op ’n kanaal in die Goulburn opvangsgebied in Australië uitgevoer. Die resultate het die belowende rol van masjienleertoepassings op die invul van gapings in stroomvloeistate getoon en aangedui dat data van meetstasies ’n groter bydrae tot die sukses van hierdie modelle lewer in vergelyking met reënvaldata. | af_ZA |
| dc.identifier.uri | http://hdl.handle.net/10019.1/103394 | |
| dc.language.iso | en_ZA | en_ZA |
| dc.publisher | Stellenbosch : Stellenbosch University | en_ZA |
| dc.rights.holder | Stellenbosch University | en_ZA |
| dc.subject | Hydrology -- Data processing | en_ZA |
| dc.subject | Streamflow | en_ZA |
| dc.subject | Machine learning -- Models | en_ZA |
| dc.subject | Rivers -- Australia | en_ZA |
| dc.subject | Forecasting -- Mathematical models | en_ZA |
| dc.title | Short-term stream flow forecasting and downstream gap infilling using machine learning techniques | en_ZA |
| dc.type | Thesis | en_ZA |