موقع ومنتديات ابن الإسلام بإشراف الأستاذ بشير بن نعمان دحان: تقييم الأراضى المستصلحة فى محافظة كفر الشيخ باستخدام الاستشعار عن بعد و نظم المعلومات الجغرافية

تقييم الأراضى المستصلحة فى محافظة كفر الشيخ
باستخدام الاستشعار عن بعد و نظم المعلومات
الجغرافية
================================================================================
إبن الإسلام on 10/01/2012 18:57:00

منى فيصل محمد  عين شمس  هندسة      الهندسة
البيئية          دكتوراة    2005

Assessment of Reclaimed Land in Kafr El-Sheikh Governorate Using Remote Sensing
and GIS"
This study models soil salinity in the north of the Nile Delta by using
Geographic Information Systems (GIS) and remote sensing. The study area is
located in El-Hamul area to the south-east of Lake Burullus, which mostly
suffers from water logging and high salinity levels. A detailed survey was
carried out on a grid system with 2.5 km lag interval, where soil samples and
groundwater samples were tested for the main chemical and physical
characteristics.  Point maps with attributes were created on ArcGis 9.0 and
interpolations were made using Spherical Kriging Models. Two landsat TM images
dated June 1988 and June 1998 used for the study with the aid of ERDAS Imagine
8.5. The Normalized Difference Vegetation Index (NDVI) has been calculated to
separate bare and vegetated areas. The vegetated and bare areas were used as a
base to separate the corresponding area at different bands such as salinity,
texture and groundwater for both vegetated and bare areas.  The digital values
of the tested layers were input to Neuralyst v1.4 Software to use the Artificial
Neural Network (ANN) technique to create two networks (one for bare and the
other for vegetated soil) and to estimate the soil salinity.  Neuralyst v1.4 was
applied on two different applications for vegetated and for bare soils. To
design, implement and validate ANN technique in soil salinity detection, the
first stage comprised the adaptation of the network parameters for optimizing
stability and achieving minimum output error, and the second stage was to
validate the results. The resulting calibrated model was applied on the image
data of 1998 to establish a new salinity map. The resulted salinity map was
verified with field samples at 145 points. In the regression analysis a positive
relation between the measured and the predicted soil salinity was determined
R2=0.72, To point out the differences between the soil salinity of year 88 and
98, most of the study area is degradation representing 51 % of the total area
while the constant and improved area represented 22 % and"