03448nas a2200265 a 450000500170000000800410001703500150005810000230007324501110009626000670020730000230027449000270029750001550032450200580047952021400053765000580267765000620273570000360279770000540283370000470288771000830293471000400301781000590305785600660311620260817163441.0170918s2015    th uu m rtt   000 a|eng d  a.b121916190 aSasirin Srisomkiew10aAssessing soil salinity using the relationship of satellite image observation with electrical conductivity  aPathum Thani, Thailand :bAsian Institute of Technology,c2015  a82 leaves :bill. 1 aThesis ;vno. RS-15-02  aA thesis submitted in partial fulfillment of the requirements for the degree of Master of Science in Remote Sensing and Geographic Information Systems  aThesis (M. Sc.) - Asian Institute of Technology, 2015  aThe problem of soil salinity in Thailand is prominent, especially in the Northeast region. Remote sensing technologies have recently been employed for mapping soil salinity. The objective of this study is to develop a technique of utilizing satellite remote sensing data for assessing soil salinity. Five districts from the province of Nakhon Ratchasima were selected for this study from which 30 different soil samples were collected for laboratory analysis. The correlation of spectral reflectance with electrical conductivity was established by using the remote sensing data from Landsat 8 OLI and laboratory EC. Analysis showed blue band, NDSI and salinity index S4 have high correlation with observed EC. The multiple regression analysis between EC and the spectral reflectance generated the 8 models which showed R2 more than 70%. Using the regression equation from  model  generated  we  predict  the  EC  value  for  soil  samples  and  each  pixel  of Landsat 8 OLI data. Subsequently the soil salinity map was generated by  classifying EC according to its observed value.  Verification of models from Landsat 8 OLI was done using HJ-1A satellite image to check if models be used with other satellite. Regression equation of model 7 and 8 of Landsat 8 OLI was used to predict EC for HJ-1A as all band of Landsat 8 OLI was not available. The soil salinity map was then generated for HJ-1A for which the accuracy of the salinity map was evaluated by comparing random pixel from model 7 and model 8 of Landsat 8 OLI and HJ-1A with the soil series data. The overall accuracy of 65% and 70% was observed for model 7 of HJ and model 7 of Landsat 8 OLI with the soil series data. Outlining the overall study, the correlation analysis between prediction variables and observed EC was useful to study depth on the prediction model and hence generate the soil salinity map. The map thus provided us information of distribution of salinity in the study  area.  The  acceptable  accuracy  from  accuracy  assessment  thus  implied  that prediction model and map produced was well-suited for mapping and monitoring the salinity in the area 0aSalinization, Remote sensingzThailand, Northeastern  0aSoil salinizationzRemote sensingzThailand, Northeastern0 aSarawut Ninsawat, eChairperson1 aShrestha, Rajendra Prasad,eExamination Committee1 aNakamura, Shinichi,eExamination Committee2 aMinistry of Agriculture and Cooperatives (MOAC), Thailand, eScholarship donor2 aAIT Fellowship, eScholarship donor2 aAsian Institute of Technology.tThesis ;vno. RS-15-02  3Full-Textuhttp://203.159.5.9/ait-thesis/detail.php?q=B03906