SPECTRAL SIGNATURE IDENTIFICATION AND SURFACE AREA ESTIMATION FOR FRUIT CROPS GROWN IN TIBA REGION OF WESTERN NILE DELTA OF EGYPT BY USING LANDSAT-8 SATELLITE REMOTE SENSING DATA | ||||
| Journal of Plant Production | ||||
| Article 1, Volume 6, Issue 12, December 2015, Page 1917-1940 PDF (1.24 MB) | ||||
| Document Type: Original Article | ||||
| DOI: 10.21608/jpp.2015.52120 | ||||
 View on SCiNiTO | ||||
| Author | ||||
D. O. El-Ansary 
| ||||
| Department of Pomology, Faculty of Agriculture (El-Shatby), University of Alexandria, 21545, Egypt | ||||
| Abstract | ||||
| In this research, satellite remote sensing data and field measured ground-truth data were used to identify the spectral signature and estimate the total surface area for fruit crops grown in Tiba region of western Nile delta of Egypt. Global position system (GPS) data collection and surface bare soil sampling for this research was conducted in Tiba region during 2012 and 2013 seasons. Ground-truth data included: eleven fruit crops: orange, mandarin, lime & lemon, grape, apple, peach, apricot, plum, guava, mango, and cactus pear; six other crops: tomato, squash, wheat, berseem clover, broad bean, and sugarbeet; and finally bare soil. One Landsat-8 Operational Land Imager (OLI) and Thermal Infrared Sensor (TIRS) satellite image for the study area covering WRS-path 177 and WRS-row 39, acquired on 19 March 2013 was radiometric calibrated to top of atmospheric reflectance and spectral signature curves were developed for each crop and for grouped crops against bare soil. Nine vegetation indices (VIs) were derived from the reflectance wavelength of the visible, near infrared and shortwave infrared parts of spectrum and included: Normalized Difference Vegetation Index (NDVI), Ratio Vegetation Index (RVI), Transformed Vegetation Index (TVI), Ashburn Vegetation Index (AVI), Soil-Adjusted Vegetation Index (SAVI), Modified Soil-Adjusted Vegetation Index 2 (MSAVI2), Tasseled Cap Brightness (TCB), Tasseled Cap Greenness (TCG), and Tasseled Cap Wetness (TCW). Supervised classifications for band reflectance and VIs were performed to estimate areas of fruit crops, other crops and bare soil, and accuracy assessment for the developed research methodology was presented. Moreover, multiple linear regression modeling (MLRM) equations to predict bare soil calcium carbonate (CC) %, electrical conductivity (EC) in dS·m-1, pH, and soil texture (sand, silt, and clay %) by using the first eight bands of Landsat-8 OLI remote sensing data of bare soil as regressors were also developed, and soil mapping units (SMU) in Tiba were presented by using geographic information system (GIS) interpolated maps. Results indicated that satellite band reflectance estimates for cropping pattern were more precise than VIs, and it is recommend to use reflectance of band 3 (0.53 – 0.59 µm) or 4 (0.64 – 0.67 µm) to estimate surface area for fruit crops. MLRM prediction indicated that in most Tiba region CC, EC, and SMU of bare soil were predicted with high accuracy. | ||||
| Keywords | ||||
| Remote Sensing; Landsat-8; Spectral signature; Vegetation Indices; cropping pattern; fruit crops; soil mapping units | ||||
|
Statistics Article View: 168 PDF Download: 731 |
||||
