The relationship between soil salinization and various water management practices at large scales has not been fully revealed due to the difficulty in obtaining spatial and temporal information on soil salinity and water management. Therefore, obtaining accurate spatiotemporal information on soil salinity and water management simultaneously is key to unravel the relationship.

In a paper published in Science Bulletin, a research group led by Prof. LUO Geping at the Xinjiang Institute of Ecology and Geography of the Chinese Academy of Sciences presented evidence of heterogeneous water management altering dryland’s soil salinization and degradation neutralization.

They found that advanced practices, such as drip irrigation and pipe drainage, can substantially reduce salinization, compared with conventional practices, such as flooding irrigation.

The research team combined the large number of field samples collected from Xinjiang, Central Asia, and the Middle East with the salinity samples from the World Soil Information Service database and estimated the topsoil salinity of dryland cropland at 1km resolution from 1984 to 2021.

"The salinity estimation was based on machine learning and Google Earth Engine. These new technologies can effectively support the accurate global-scale soil salinity mapping," said SHI Haiyang, the first author of the study.

In addition, the output of the advanced high-resolution global water use model PCR-Globwb was used in this study to provide more reliable information related to water use and groundwater management. Remote sensing-based information on soil plastic mulch and machine learning-based information on the distribution of tile drainage were also used, fully utilizing the advantages of remote sensing and machine learning in obtaining information on water management practices at large scales.

"Based on remote sensing and machine learning, the feasibility of obtaining many large-scale irrigation and drainage information that was difficult to obtain in the past has been greatly improved,” LUO Geping said.

In the future, the combination of new technologies such as machine learning, remote sensing big data analysis, and high-resolution earth system models will advance the research on the relationship between soil salinization and water management to a greater extent.（from XIEG）

Article Link: https://www.sciencedirect.com/science/article/pii/S2095927323007661?via%3Dihub