Spatial Discrimination and Mapping of Soil Salinity Using Remote and Gis Techniques in the Middle Awash Basin
No Thumbnail Available
Date
2007-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Addis Ababa Universty
Abstract
This research deals with the identification and mapping of salt affected soils in
Middle Awash. The analysis was made based on remote sensing data of Landsat
ETM+ image, Electrical conductivity (EC) of soil sample taken in the same year when
the image was captured and some other ancillary data. Remote sensing techniques
employed were visual interpretation, digital image processing and indices like
Normalized Difference Salinity Index (NDSI) and Normalized Difference Vegetation
Index (NDVI).
An attempt has been made to identify salt affected area by visual interpretation using
both TCC and FCC. From the unsupervised and supervised classification1930ha and
887.3ha areas have been mapped as highly saline. The result obtained from
unsupervised classification is larger in area extent. However the result obtained from
NDSI was not only in area wise but also the level of salinity as highly saline,
moderately and slightly saline, determined based on the reflectance value. Out of the
total area 10% are mapped as highly saline. Attempt has been made to indirectly
identify salt affected area using NDVI and the result showed similar area distribution
of salt affected area.
A regression analysis between EC values of small areas confined only in Melka Sadi
area and the corresponding reflectance value in the NDSI image offer a polynomial
relation of order two. The model obtained used to derive a salinity map and 16 and
0.9 dS/m of maximum and minimum EC level respectively was estimated. The spatial
distribution of salt affected area derived from NDSI and model has shown similar
pattern but different area extent.
Overlay analysis of salt affected area and soil, canal, drainage and water table has
been made to assess the spatial distribution as well as the relationship with this
features and the result revealed the spatial distribution not highly influenced by the
features being considered.
Key word: NDSI, EC, Soil salinity
Description
Keywords
NDSI, EC, Soil salinity