Soil is part and parcel of operating essential functions of ecosystems. Its degradation including salinization may affect negatively both agriculture and human life. Soil salinization is a form of environmental destruction resulting in loss of soil quality, especially in dry and semi-arid zones including both primary and secondary salinization. Today, salt-affected areas cover 7% of the Earth’s surface with hotspots in the United States, Argentina, Pakistan, China, Sudan, and many states in Central and Western Asia existing in a minimum of 100 countries. Timely and effective management is required to preserve soil fertility for the next generation. Currently, modeling is an effective approach to identify, evaluate and predict soil salinization distribution. Modeling with remote sensing techniques is widely used to identify salt-affected soil, map saline locations, and estimate salinity levels. Today, a lot of geostatistical techniques are applied to predict soil salinity over time including Ordinary Kriging, Co-kriging and Indicator Kriging. The current study originated as a review with the purpose of exploring the scientific background of soil salinity modeling throughout the world. Our research, which includes 177 Scopus-indexed final published articles in English, focuses at publishing trends, types, top publishers, notable authors, affiliated institutions, geographical and thematic emphasis areas, co-authorship and keyword occurrences. Additionally, this paper provides a review of the most popular models and conditioning factors. Among various models, multiple linear regression was used as the most prevalent model type for soil salinity estimation in 10.1% of the publications, and electrical conductivity is considered as the most value conditioning factor in 19.2% of analyzed papers. Finally, it is concluded that recent technological developments in remote sensing are significant factors in increasing research interest in soil salinity modeling.
Read full abstract