Estimation of salinity and heavy metals over marshlads based on landsat-8 data
Loading...
Date
2017
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Universiti Teknologi Malaysia
Abstract
Marshes are the provider of habitat for several types of living creatures. Their preservation are prioritized for sustainable environment and eco-friendliness. Iraqi marshland is the largest wetland with an area of 15,000-20,000 km2 in the Middle East and Western Eurasia and has a significant impact on the ecosystem. The salinity in the Tigris and Euphrates Rivers near their discharge point at the marsh ranges from 0.5 to 2ppt (parts per thousand). This thesis focuses on Al-Hawizeh marsh, which is one of the major marshes with an area of 2,500-3,000 km2 in Iraq and considered as an enriched resource of fishing and irrigation. Of this mashland, 74% of it’s agricultural land suffers from high degree of salination that need to be overcomed. Several man-made activities and post-war related events have caused radical deterioration of water quality in this marshland. The aim of this study is to monitor and assess the water quality parameters of this marsh. The optical remote sensing dataset (bands B6, B7, and B11) from Landsat-8 (OLI/TIRS) are synergistically integrated to the proposed salinity index (SI) and soil moisture index (SMI) model. By using the newly developed algorithms, the optimum water quality parameters in terms of salinity and minerals contents which comprised of iron, lead, zinc, nickel, calcium carbonate and sulphate are determined. This creative integration between remote sensing data and developed algorithms is established to successfully map the spatial variation of salinity and minerals distributions within Al-Hawizeh marsh during four seasons in the year 2013. The results of this study show that SMI model achieved better accuracy in retrieving the water quality parameters than the SI model. The average of the concentrations values for (salinity, SO4, CaCO3, Fe, Pb, Ni and Zn) by using SMI model are found to be minimal in winter as (746, 121, 84, 0.59, 0.49, 0.04 and 0.036) mg respectively and maximum in autumn as (1956, 202, 172, 0.64, 0.53, 0.08 and 0.05) mg respectively. The decision tree (DT) classification that uses single band outperformed the support vector machine (SVM) classification when combined with SMI model. This study also found that the change of value for salinity and mineral are minimum between winter and spring but maximum between summer and autumn. In conclusion, the developed systematic and generic approach may constitute a basis for determining the water quality parameters in the marshland worldwide
Description
Thesis (Ph.D (Geomatics Engineering))
Keywords
Geoinformation and real estate, Remote sensing