How much down the ditch ? An assessment of N in River La Chaux catchment of Mauritius
| MSI07P4229 | |
| Chung Tze Cheong, M. J. S. Y. | |
| School of Geography, Planning and Archicteture, University of Queensland, Australia | |
| How much down the ditch ? An assessment of N in River La Chaux catchment of Mauritius | |
| Thesis, Master of Philosophy | |
| Thesis | |
| 2006 | |
| 200 p.: 23 tbls, 42 figs, 8 pl, 4 appen. | |
| En | |
| En | |
| Every year 11,000 tonnes of N is being applied to the sugarcane fields in Mauritius, which occupy 43 percent of total land area. The perception that sugarcane production is a potential pollutant to the rivers and the lagoon is felt strongly. Though the findings from samples collected in River La Chaux showed that the N detected was not of agronomical significance, and was just within the threshold value for environmental concerns, refinement to these initial findings was necessary to verify possible undetected peak values, which were not revealed by the simplicity of a lumped model. The objectives of this study are: To define a conceptual model for River La Chaux catchment hydrology, taking into account the water balance, and hence the transport of different forms of N within the catchment's sources and sinks; To identify the sources and sinks of N, the key parameters which influence their transport in the catchment, through the integration of surface and subsurface hydrology model, where a conventional hydrology model would be limiting because of the complex volcanic geology of the island. The project study focus is to understand the processes of catchment hydrology and nutrient transport in River La Chaux catchment through a conceptual model, and to translate it to a catchment model that corresponds most to the conceptual model. Soil and Water Assessment Tool (SWAT) was chosen and tested for this simulation exercise. It is a physically distributed catchment model that caters for surface and subsurface hydrology interaction, including simulation of crop growth, the transformation processes of N nutrient, and its transport within the catchment. GIS processing facilities were used to define the sub-basins of the catchment into hydrologic response units (HRU), based on topography, land use, and soil type characteristics. The defined HRUs were coupled with the daily time series meteorological data to establish the hydrology components of the flow. This rainfall-runoff model enables simultaneous calculation of lateral and vertical flow through soil profile, in addition to the surface runoff. A commercial crop calendar was also linked to the model to represent the standard cropping practice of cane growers in terms of planting and harvest dates, land preparation, N input and expected crop yield. Since the sugarcane crop occupies 87 percent of the catchment, the hydrology model hinges on realistic crop growth simulation in terms of water and nutrient uptake by the crop, and its subsequent influence to other hydrology components like run-off, lateral flow and percolation. Modifications to parameters in SWAT crop database with local values and recent findings in the sugarcane crop modeling improved the simulated biomass and evaporation values considerably. With the problem of incomplete climate datasets, two methods for potential evapotranspiration calculation were tested. The Hargreave method was found to give better potential evapotranspiration (PET) but lower biomass estimate than that of Penman-Monteith method, as it uses only the two predominant parameters: solar radiation and temperature, reducing parameters errors. However, for longer time series simulation Penman-Monteith method performed slightly better, giving satisfactory evapotranspiration and biomass simulation. Calibration of the catchment hydrology was done, using historical data of rainfall-runoff records, and the hydrology components were evaluated for the catchment. The results compared well with those obtained from the lumped model, Hydrology Simulation Program-Fortran (HSPF). It was found that rainfall in the catchment contributed to 14 percent runoff, 16 percent lateral flow, and 37 percent percolation, 31 percent of evapotranspiration, and 2 percent as deep drainage loss. The surface run-off however was much more important in SWAT, and insignificant the deep aquifer loss, as opposed to the findings of HSPF. Comparison of simulated flow with measured flow, using the Nash-Sutcliffe model efficiency evaluation method, gave values between - 0.39 to 0.14, and the R2 value 0.7 - 0.8, with a systematic overestimation of 30 to 40 percent of measured flow. Dilution of the model's efficiency was exhibited with long term historical data, indicating inherent measurement errors in these data. The model also showed that the upstream catchment boundary does not reflect the groundwater boundary. However despite the unaccounted transmission loss in geological faults and collapsed lava tunnel in the catchment, the majority of the fast moving base flow exits at River La Chaux estuary. The study highlighted the preponderant role of sugarcane crop in the hydrology of the catchment in attenuating run-off, and in ensuring the replenishment of the aquifer. Nutrient transport is mainly by base and lateral flows, which prevail even in dry period. Detailed water balance for each sub-basin revealed the difference in N transport pathways in the sub-basins: Percolation predominates in young lava soils, whereas lateral flow in intermediate lava soils. Erosion is not a problem in the catchment. Simulation results ranked the agronomic importance of N loss as follow: Leached N <17.5 percent, lateral flow N < 15 percent, and surface runoff N < percent of applied N fertilizer. Though the simulated N budget for sugarcane cropping system reproduced relevant N transformation rates, and showed the N sources and sinks in a predominantly sugarcane growing catchment. Insufficient representation of N immobilization by the model was noted for long cycle crop like sugarcane. The findings showed that 'wet season' nitrate-N loss could be quite high in sub-basins where lateral flow and preferential pathways like drains were found. During the 'dry' season, nitrate transport is negligible. From the 'snapshot' sampling, insignificant N contribution of sugar industry to stream water is confirmed, whereas other activities done at the expense of the riparian vegetation seemed to increase N pollution considerably in the downstream area even during 'dry season'. However, the hot spots of point source and non-point source N contribution in the catchment went undetected at next sampling point, as the nitrate was well assimilated by the stream, thus pointing to the necessity of research into 'untouched areas' like stream ecology and the role of riparian zone, to complete the picture of catchment N dynamics in the future. The model showed how much gone down the ditch, but did not clarify the fate of N in the ditch. On the whole the objectives have been achieved. Better understanding of the flow and N transport pathways of this geologic complex catchment was allowed with the study. The underlying conceptual model of SWAT has been tested for non-point source pollution of the catchment, and can be adapted to Mauritian catchment for hydrology modeling | |
| water quality water catchments hydrology models modelling water balance Nitrogen nitrogen, sources of nitrogen, sinks of nitrogen transport pathways River La Chaux sugarcane Nitrogen pollution water pollution | |
| Mauritius | |
| Water use and management | |
| Water pollution | |
| 2007-01-11 | |
| En | |
| LIB CHEM | |
| CAT | |
| CHEM |