Effects of urbanisation in Upper Manyame on the flow of Manyame River at Gauging Station C21

Abstract

Regression analysis was used to establish the relationship between urbanisation in Harare and river yield of Manyame River at gauging station C21. To detect the effects of urbanisation, Landsat images were downloaded from USGS website (http://landsatlook.usgs.gov/viewer) and they were processed in Quantum GIS and System for Automated Geoscientific Analysis. Image classification was carried out in SAGA were the images were classified into five main classes which were Forest, Scrub, Ground, Water and Built up. Change detection was then carried out in SAGA and also the changes that occurred were calculated as a percentage in SAGA. The results showed that in Upper Manyame, forests were totally deforested by 18.04 percent. The area that was covered by scrub was reduced by 10.2 percent. Built up area increased sharply by 11.2 percent and the area that was covered by bare ground drastically expanded by 17.8 percent. The average NDVI for C21which showed the reduction in vegetation was calculated in Arc GIS. The median NDVI values were then calculated in Excel and it showed that there was a general decrease in the NDVI values in C21 catchment with the highest 0.47 in 1998 showing a healthier forest and lowest was 0.36 in 2011 showing a degraded forest. The river yield for Manyame river at gauging station C21 was calculated in excel and showed that there was a general increase in the river yield as a percentage from 104 percent to 129 percentage although there was a sharp decrease in 2008 to 109 percent where there was a draught in Zimbabwe. Linear least square regression analysis was run to establish the relationship between river yield and NDVI and the relationship was statistically significant with p<0.05 tested by Gen stat 2014 edition . The NDVI and river yield had a strong inverse relationship of -0.72. The r2 value for NDVI and river yield was 0.5199.