Search

Preventing soil salinization is key to the healthy development of agroecosystems in arid regions. Notably, long‐term anthropogenic irrigation accelerates secondary salinity accumulation in arid areas; however, the regional‐scale spatial patterns of soil salinity pre‐ and post‐irrigation season remain unclear. Accordingly, field observations over two periods (n = 182 samples), along with environmental covariates and the randomForest algorithm, were used to estimate the horizontal and vertical spatial distributions (12 soil depths) of soil salinity in the Weiku Oasis (0–60 cm) at a 30 m spatial resolution. The results showed that the prediction accuracy was greater for the non‐irrigated season than for the irrigated season. Moreover, soil salinity was low in the interior of the oasis, high in the desert oasis interlacing zone, and significantly decreased with increasing soil depth. Variable contributions indicated that the effects of SoilGrids variables on soil salinity increased with greater soil depth. During the irrigation season, an inflection point for salt aggregation was observed at ~55 cm for farmlands, grasslands, and shrubs, with the greatest variation seen in farmlands. The findings here provide insights into the improvement of irrigation techniques in arid zone agroecosystems. [ABSTRACT FROM AUTHOR]

• GMM can distinguish similar spectral features in vegetation. • After spectral feature partition, the difference modeling is superior to whole-sample modeling. • UAV hyperspectral can indirectly reveal salinization patterns by inversion of saline vegetation SPAD. The chlorophyll content of saline vegetation can indirectly reflect salinization. Rapid and non-destructive capture of chlorophyll content of saline vegetation at a regional scale is essential for saline soil improvement and sustainability of saline agriculture. However, traditional soil–plant analyzer development (SPAD) monitoring based on SPAD-502 is carried out at the leaf scale, which does not allow rapid access to SPAD information for the whole region. In this study, we proposed that post-hyperspectral classification based on spectral differences that could contribute to enhanced estimation of SPAD in saline vegetation. To test this proposal, we partitioned the hyperspectral images using a Gaussian mixture model. Then, estimation models based on spectral partition and full-sample SPAD were developed using in situ observation data and a random forest model, respectively. Finally, the unmanned aerial vehicles (UAV) hyperspectral images were used as the input data source to digitally map the SPAD of saline vegetation in the region, using the prediction model. The results indicated that there were significant intensity and shape differences in the spectral reflectance characteristics of saline vegetation under different clusters. The SPAD prediction model, based on spectral feature partition, performed significantly better than the full sample. The SPAD maps of saline vegetation before and after clustering displayed similar spatial distribution models, but the prediction uncertainty of the models, based on spectral feature partition, was relatively low. Our results confirm the effectiveness and stability of UAV hyperspectral and spectral partition-based modeling in developing SPAD spatial distribution estimation models for saline vegetation. [ABSTRACT FROM AUTHOR]

[Display omitted] • We updated information on soil salinity: soil salinization exacerbated overall. • Spectral information and environmental covariates were necessary but still needed to filter. • The model with environmental covariates improved on performance for soil salinity. • The GBRT model performance was the best among machine learning models. Soil salinization is the main source of global soil degradation. It has impeded progress towards sustainable development goals (SDGs) by threatening 20% of irrigated areas. However, in many data-poor regions, accurate soil salinization information is unavailable. Thus, an updated soil salinity map with high accuracy and resolution is urgently needed to help local governments conduct precise management. In this study, a bootstrap hybrid machine learning framework was developed combine Sentinel-2 data and environmental covariates. The Boruta algorithm was applied to input the spectral information and environmental variables to identify the primary factors influencing soil salinity. By averaging 100 model iterations within a bootstrap framework, the soil salinity mapping outcomes were compared from four machine learning methods (bagging, classification and regression tree, random forest, and gradient boosting regression tree (GBRT)). The results showed that the models driven by spectral information and environmental covariates (strategy II) explained 68∼88% of the variability in soil salinity. The model accuracy of strategy II was improved by 5%-8% over that of the models driven only by spectral information (strategy I). The GBRT yielded the most appropriate averaging performance of the four machine learning approaches within strategy II, with an R2 of 0.88, a root mean square error (RMSE) of 6.33 dS m−1, a ratio of performance to interquartile distance (RPIQ) of 4.66 and model stability (ROB) of 0.44. The bootstrap averaging method was relatively stable and had high accuracy potential. The distribution of soil salinity in the Ebinur Lake region was the result of a combination of natural and human activity influences. The proposed approach provides a soil salinity mapping strategy with a fine resolution (10 m) and high accuracy in data-poor places. It may also aid in the restoration of biodiversity, the decrease in land degradation, and the avoidance of future food output reductions in the future. [ABSTRACT FROM AUTHOR]

• Successful application of the ESTARFM to study the dynamics of playas surface. • NDVI and SI are sensitive to precipitation events. • Precipitation events affect the spatiotemporal distribution of surface salinity. • Salinity spectral indices decrease after precipitation events. Playas are desert landscapes unique to arid regions that respond quickly to climate change. Changes in surface salinity in Praia are directly linked to water security and regional ecological and economic health. This study aimed to explore the sensitivity of remotely sensed spectral indices to precipitation events. An ESTARFM was utilized to fuse daily (less than 10% cloudiness) MODIS and Landsat 8 imagery from April to September 2019 in the Ebinur Lake Wetland Reserve. Salinity was tested in the laboratory after soil collections in May (spring) and August (summer). The significantly correlated normalized difference vegetation and salinity indices were compared to the precipitation data and found to fluctuate in response to precipitation. The normalized difference vegetation and salinity indices decreased following precipitation events and increased during precipitation intervals, and the precipitation amount and evaporation intensities may have influenced the magnitudes of the decrease and increase. Following precipitation, the playa surface produces new puddles and a new spatial distribution pattern of soil salinity. This study determined that the spatiotemporal fusion technique is an effective method for observing the dynamics of playa surfaces, and precipitation and evaporation affect the spatial distribution of salt on the playa surface; however, the monitoring period should be short when utilizing remote sensing to monitor playa salinity. [ABSTRACT FROM AUTHOR]