Estimating soil salinity with different fractional vegetation cover using remote sensing

• DOI: 10.1002/ldr.3737
• Publication Date: 2020-08-09T16:40:00Z
• Authors: Junrui Zhang; Zhitao Zhang; Junying Chen; Haiying Chen; Jiming Jin; Jia Han; Xintao Wang; Zhishuang Song; Guangfei Wei
• Keywords: 2. Zero hunger; 0401 agriculture, forestry, and fisheries; 04 agricultural and veterinary sciences; 15. Life on land

Abstract Soil salinization is a serious restrictive factor affecting sustainable agricultural development. In order to explore the effect of Fractional Vegetation Cover (FVC), we monitored soil in sites different vegetation coverage Jiefangzha Irrigation District Inner Mongolia using satellite remote sensing. From May August 2018, carried out field sampling at depths each month, and calculated FVC spectral covariates GF‐1 images corresponding period. Based on division criteria for Mongolia, took following steps: (a) setting up control treatment A (the full data with undivided FVC, TA) experimental treatments B (bare land, TB), C (mid‐low TC), D (mid TD) E (high TE); (b) conducting Best Subset Selection (BSS) all treatment; (c) constructing Salt Content (SSC) inversion models partial least square regression (PLSR), Cubist, Extreme Learning Machine (ELM). The results indicated that classifying could improve stability predictive ability models; performance three modeling methods were (Cubist was best, ELM next PLSR poorest); optimal TB, TC TE constructed by Cubist 0–20, 0–40 0–20 cm, TD 0–60 respectively. can provide references prevention production other areas similar cover.