Thesis

 

 

Abstract:

110 years of intensive mining has taken place on the copper belt province in Zambia. The copper boom of the 60′s, 70′s and late 2000′s has seen new and massive expansion in mining projects of both surface and underground mining operations in Chingola town. The subsequent influx of people, the growing levels of poverty among the people and the increase of informal settlements around the town have all contributed to the current deteriorating environmental and social conditions. Remote sensing is used in this study to analyze the land change in the town and its surrounding area by using images from Landsat MSS in 1972 (4 bands), ETM 1989 (6 bands) and ETM 2002 (6 bands) in addition the Topographic map of 1960 and Google Earth were used as reference base maps of the Town.

ERDAS software is used in classifying the images. Supervised classification is applied so as to detect the changes in Land cover-use pattern over a period of 30 years. With the use of maximum likelihood supervised classification system various land-use classes; woodland, grassland, cultivated land, bare soil, rivers, dams, water ponds, built-up area and open cast mines were identified from satellite data and Google Earth surveys. LandScape metrics are calculated using Fragstat and visualised in ArcGIS,

IDRISI software is used in the post-classification change detection method it gives a very effective method in detecting land-use change in the various images. Landsat ETM+ 2002 was used as the reference image. MSS 1972 (winter season) was compared with Landsat ETM 2002 (winter season) to determine how the change are progressing in the study period.1972 is then compared with the 1989 images to look at the changes. There have been significant and extensive changes in the LULC cover patterns in the last 30 years. Change detection images highlight that bare ground, agriculture areas and built up areas have increased whereas water bodies and the Miomba woodland has undergone a massive decrease; woodland and grassland have been changed to cultivated and fallow land. A few tree/wood plantations were established over the last 30 years.

 

 

Figures above display the statistic values in form of pie charts against the classified maps during the study period. The study revealed that in 1972 the study area was mostly covered with the Miomba woodland (65%) which by 1989 was reduced to 39% and further went down to 30% by the year 2002. Agriculture area on the other hand showed an increase from 26% in 1972 to 52 % in 1989 and reached a level of 58% in 2002. Bare ground showed a steady increase from 1% in 1972 to 3% in 1989 and consequently 4% in 2002. Water bodies were reduced from 4% in 1972 to 1% in 2002.Built up areas showed an increase from 4% in 1972 to 7% in 2002.

 

 

 

 

 

 

 

 

Grid square based analysis is applied in making a density change analysis between the study periods. This GIS methodology uses a point in polygon by first cutting the study area into equal size analysis squares, in this case 100 by 100 and then calculating the sum of points in each analysis square polygons. This technique gives a clear visual interpretation of the changes that has taking place.