We use cookies to ensure that we give you the best experience on our website. By continuing to browse this repository, you give consent for essential cookies to be used. You can read more about our Privacy and Cookie Policy.

Durham e-Theses
You are in:

Vegetation change detection and soil erosion risk
assessment modelling in the Man River basin,
Central India

THAKUR, JITENDRA (2015) Vegetation change detection and soil erosion risk
assessment modelling in the Man River basin,
Central India.
Doctoral thesis, Durham University.

PDF (Vegetation change detection and soil erosion risk assessment modelling in the Man River basin, Central India) - Accepted Version


Land use change directly increased soil erosion risk, which is a very sensitive environmental
issue in Central India. To evaluate the response of land use changes on soil erosion risk,
research was implemented using remote sensing techniques, coupled with ground
information, to develop an integrated modelling approach to study the factors driving land
use changes in the Man River basin, Central India. Results were used to assess the impact of
land use change on soil erosion risk.
First, a series of sub methods were applied to monitor and verify land use land cover change
in the study area which included pre-processing, classification and assessment of land use
transaction from 1971 to 2013 using Landsat time series imagery. Additionally, an
independent spatial assessment of deforestation, forest degradation and responsible drivers
for the period 2009-2013 was conducted to enable a deeper analysis of forestry activates
using the GIS based direct interpretation approach. The research also developed a robust
accuracy assessment method to check the quality of the 2009 and 2013 classification maps
using good quality Google Earth TM imagery and a field measured GPS dataset. These
approaches were largely based on the GOFC- GOLD (2010) and IPCC good
recommendations for land use land cover mapping and verification. The information
obtained from an accuracy assessment was also used to estimate deforestation area and
construct confidence intervals that reflect the uncertainty of the area estimates obtained.
Such analysis is rarely applied in current published verification assessments.
In the second phase of the study, a Geo-spatial interface for process-based Water Erosion
Prediction Project (GeoWEPP) was implemented, to estimate the response of land use and
land cover change on soil erosion risk in several scenarios derived from both ground and
satellite based precipitation, DEMs and vegetation change. GeoWEPP was used at the
hillslope scale in three selected watersheds within the Man River basin using Landsat, LISSIII,
Cartosat-1, ASTER, SRTM, TRMM and ground based datasets.
The results highlight that the study developed a realistic approach using remote sensing
techniques to understand the pattern and process of landscape change in the Man River basin
and its response on soil erosion risk. Over the last four decades, forest and agriculture areas
were found to be the most dynamic land use /land cover categories. During the last four
decades, around 54200 ha (33.7 %) forest area has been decreased due to the expansion of
agriculture, forest harvesting and infrastructure development. The direct interpretation
approach estimated similar patterns of deforestation and forest degradation associated with
drivers for the 2009 to 2013 time period, but this approach also provided more accurate and
location specific information than automatic analysis. The overall correspondence between
the map and reference data are a good measure for 2009 and 2013; 94.03 % and 92.8 %
respectively. User‘s and producer‘s accuracies of individual classes range from 75 % to 99
%. Using the accuracy assessment data and a simple set of equations, an error-adjusted
estimate of the area of deforestation was obtained (± 95% confidence interval) of 23382 ±
550 ha.
The estimated average annual soil loss for all three watersheds is 21 T/ha which was found
to be comparable to similar studies carried out in the study region. The highest soil loss rates
occurred in areas of agriculture (301 T. /ha /yr) and fallow land (158 T/ha/yr), while the
lowest rates were recorded in forest land (33.45 T/ha/yr). Agriculture extension (316.5 ha)
due to forest harvesting (234 ha) in the last four decades is one of the significant drivers to
speed up soil erosion (7.37 T/ha/yr.) in all three watersheds. The spatial pattern of erosion
risk indicates that areas with forest cover have minimum rates of soil erosion, while areas
with extensive human intervention such as agriculture and fallow land, have high estimated
rates of soil erosion. The different DEMs generated varied topographic and hydrologic
attributes, which in turn led to significantly different erosion simulations. GeoWEPP using
Cartosat-1 (30 m) and SRTM (90 m) produced the most accurate estimation of soil loss
which was close to similar already published studies in the area. TRMM rainfall data has
good to use as a rainfall parameter for soil erosion risk mapping in study area.
Overall, the integrated approach using remote sensing and GIS allowed a clear
understanding of the factors that drive land use/land cover change to be developed and
enabled the impact of this change on soil erosion risk in the Man River basin, Central India
to be assessed.

Item Type:Thesis (Doctoral)
Award:Doctor of Philosophy
Keywords:Vegetation change detection, Soil erosion risk assessment,Modelling,Remote sensing techniques.
Faculty and Department:Faculty of Social Sciences and Health > Geography, Department of
Thesis Date:2015
Copyright:Copyright of this thesis is held by the author
Deposited On:06 Dec 2016 14:17

Social bookmarking: del.icio.usConnoteaBibSonomyCiteULikeFacebookTwitter