A comparative study of flow forecasting in the Humber River Basin using a deterministic hydrologic model and a dynamic regression statistical model

Picco, Robert C. (1997) A comparative study of flow forecasting in the Humber River Basin using a deterministic hydrologic model and a dynamic regression statistical model. Masters thesis, Memorial University of Newfoundland.

[English] PDF (Migrated (PDF/A Conversion) from original format: (application/pdf)) - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. Download (23MB) [English] PDF - Accepted Version Available under License - The author retains copyright ownership and moral rights in this thesis. Neither the thesis nor substantial extracts from it may be printed or otherwise reproduced without the author's permission. (Original Version)

Abstract

Since 1995 the Water Resources Management Division of the Department of Environment and Labour has generated flow forecasts for the Humber River Basin. These forecasts are required to provide a flood warning for residents living in the downstream sections of the basin. Information is also useful to the Deer Lake Power Company for the safe and efficient operation of the hydroelectric development that controls over two thirds of the basin. -- Flow forecasts for river systems can be generated using two approaches. One approach is to use a deterministic model that tries to construct a mathematical model by accounting for some, or all of, hydrologic factors responsible for runoff in the basin. The second approach is to develop a statistically based model that uses the historic flows and climate data from the basin to generate a forecast. For any model to be effective in an operational environment, it must be compatible with hydrometeorologic input data collected in near real time. -- The current method of generating forecasts uses the SSARR (Streamflow Synthesis and Reservoir Regulation) hydrologic model, a continuous simulation model that has reservoir routing capabilities, plus the ability to account for the areal distribution of meteorologic inputs, including snowmelt. This model defines the water budget using hydrometeorological inputs. -- In this study a forecast method based on the dynamic regression technique was developed to produce one, two and three day forecasts for the five gauged sub basins results. Also, forecasts for the same time periods were produced using the SSARR model. The results of both methods were compared using the mean absolute percentage error (MAPE) criterion. -- The results of the comparison showed that the dynamic regression performed better than the SSARR model for all basins, particularly for the larger basins.

Actions (login required)

View Item