To accurately forecast the future development trend of vegetation in dry areas, it is crucial to continuously monitor phenology, vegetation health indices, and vegetation drought indices over an extended period. This is because drought caused by high temperatures significantly affects vegetation. This study thoroughly investigated the spatial and temporal variations in phenological characteristics and vegetation health indices in the abdominal part of Maowusu Sandland in China over the past 20 years. Additionally, it established a linear correlation between vegetation health and temperature indices in the arid zone. To address the issue of predicting long-term trends in vegetation drought changes, we have developed a method that combines the Informer deep learning model with seasonal and Seasonal Trend decomposition using Loess (STL) and empirical mode decomposition (EMD). Additionally, we have utilized the linearly correlated indices of vegetation health and meteorological data spanning 20 years to predict the Normalized Difference Vegetation Index (NDVI) and Temperature Vegetation Dryness Index (TVDI). The study’s findings indicate that over the 20-year observation period, there was an upward trend in NDVI, accompanied by a decrease in both the frequency and severity of droughts. Additionally, the STL-EMD-Informer model successfully predicted the mean absolute percentage error (MAPE = 1.16%) of the future trend in vegetation drought changes for the next decade. This suggests that the overall health of vegetation is expected to continue improving during that time. This work examined the plant growth circumstances in dry locations from several angles and developed a complete analytical method for predicting long-term droughts. The findings provide a strong scientific basis for ecological conservation and vegetation management in arid regions.

Soil salinisation is a critical problem in northern China’s arid and semi-arid irrigated regions, posing a substantial impediment to the sustainable advancement of agriculture in these areas. This research utilises the Donghaixin Irrigation District, located on the southern bank of the Yellow River in Inner Mongolia, as a case study. This study examines the spatial distribution and determinants of soil salinisation through macro-environmental variables and micro-ion composition, integrating regression models and groundwater ion characteristics to elucidate the patterns and causes of soil salinisation systematically. The findings demonstrate that soil salinisation in the study region displays notable spatial clustering, with surface water-irrigated regions exhibiting greater salinisation levels than groundwater-irrigated areas. More than 80% of the land exhibits moderate salinity, predominantly characterised by the ions Cl⁻, HCO₃⁻, and SO₄²⁻. The hierarchy of ion concentration variation with escalating soil salinity is as follows: Na⁺ > K⁺ > SO₄²⁻ > Cl⁻ > Mg²⁺ > HCO₃⁻ + CO₃²⁻ > Ca²⁺. The susceptibility of ions to soil salinisation is ordered as follows: Ca²⁺ > Na⁺ > HCO₃⁻ + CO₃²⁻ > Mg²⁺ > K⁺ > Cl⁻ > SO₄²⁻. In contrast to the ordinary least squares (OLS) model, the geographic weighted regression (GWR) model more effectively elucidates the geographical variability of salinity, evidenced by an adjusted R² of 0.68, particularly in high-salinity regions, where it more precisely captures the trend of observed values. Ecological driving elements such as organic matter (OM), pH, groundwater depth (GD), total dissolved solids (TDS), digital elevation model (DEM), normalised difference vegetation index (NDVI), soil moisture (SM), and potential evapotranspiration (PET) govern the distribution of salinisation. In contrast, anthropogenic activities affect the extent of salinisation variation. Piper’s trilinear diagram demonstrates that Na cations mainly characterise groundwater and soil water chemistry. In areas irrigated by surface water, the concentration of SO₄²⁻ is substantially elevated and significantly affected by agricultural practises; conversely, in groundwater-irrigated regions, Cl⁻ and HCO₃⁻ are more concentrated, primarily driven by evaporation and ion exchange mechanisms.

The control of irrigation volume is of significant importance in arid regions of northwest China. Particularly, it has a crucial impact on the salinization of shallow groundwater areas. In 2022 and 2023, field experiments were conducted to test three distinct under-membrane irrigation treatments. These treatments were assigned water quotas of HW (27 mm), MW (22.5 mm), and LW (18 mm). The HYDRUS-2D model was integrated with a field experiment to accurately simulate the dynamic fluctuations of soil water and salt in the sunflower root zone. The model’s performance was assessed and verified using real-field data from 2022 and 2023, and the simulation results closely matched the measured values. This research also used stable hydroxide isotopes to assess the water supply from various soil layers at different time intervals in sunflower plants. The results indicated that the three different levels of irrigation applied under the membrane had a significant impact on soil water content. Specifically, there was a significant difference in soil water content at a depth of 0–40 cm (p < 0.05), while there was little effect on the water content at a depth of 40–60 cm (p > 0.05). After irrigation, the average salt content in the top 0–20 cm of soil decreased by 7.0% compared to the medium and low irrigation levels, and by 10.8% compared to the medium irrigation level. Additionally, the medium irrigation level resulted in a 10.8% decrease in salt content compared to the low irrigation level, and a 4.1% decrease compared to the medium irrigation level. During the same period, the soil salinity levels at depths of 0–20 cm, 20–40 cm, 40–60 cm, and 60–100 cm in the area outside the membrane were measured to be 2.7~4.8 g·kg⁻¹, 2.8~4.0 g·kg⁻¹, 2.7~3.4 g·kg⁻¹, and 1.7~2.6 g·kg⁻¹, respectively. These levels decreased by 13.1~55.5%, 0.7~42.8%, −0.4~16.2%, and −72.7~7.5%, respectively. Following irrigation, the HW treatment mostly absorbed water in the 0–40 cm soil layer, while the MW and LW treatments absorbed water in both the 0–40 cm and 60–80 cm soil levels. The results indicated that the most optimal drip irrigation method beneath the membrane in this location was achieved when the amount of water applied was between 25–30 mm. This method demonstrated a combination of water conservation, high crop yield, and effective salt suppression.

Vegetation plays a vital role in terrestrial ecosystems, and droughts driven by rising temperatures pose significant threats to vegetation health. This study investigates the evolution of vegetation drought from 2010 to 2024 and introduces a deep-learning-based forecasting model for analyzing regional spatial and temporal variations in drought. Extensive time-series remote-sensing data were utilized, and we integrated the Temperature–Vegetation Dryness Index (TVDI), Drought Severity Index (DSI), Evaporation Stress Index (ESI), and the Temperature–Vegetation–Precipitation Dryness Index (TVPDI) to develop a comprehensive methodology for extracting regional vegetation drought characteristics. To mitigate the effects of regional drought non-stationarity on predictive accuracy, we propose a coupling-enhancement strategy that combines the Whale Optimization Algorithm (WOA) with the Informer model, enabling more precise forecasting of long-term regional drought variations. Unlike conventional deep-learning models, this approach introduces rapid convergence and global search capabilities, utilizing a sparse self-attention mechanism that improves performance while reducing model complexity. The results demonstrate that: (1) compared to the traditional Transformer model, test accuracy is improved by 43%; (2) the WOA–Informer model efficiently handles multi-objective forecasting for extended time series, achieving MAE (Mean Absolute Error) ≤ 0.05, MSE (Mean Squared Error) ≤ 0.001, MSPE (Mean Squared Percentage Error) ≤ 0.01, and MAPE (Mean Absolute Percentage Error) ≤ 5%. This research provides advanced predictive tools and precise model support for long-term vegetation restoration efforts.

Drought, a significant calamity in the natural domain, has extensive worldwide repercussions. Drought, primarily characterized by reduced soil moisture (SM), presents a significant risk to both the world environment and human existence. Various drought indicators have been suggested to accurately represent the changing pattern of SM. The study examines various indices related to the Drought Severity Index (DSI), Evaporation Stress Index(ESI), Vegetation Supply Water Index(VSWI), Temperature-Vegetation Dryness Index(TVDI), Temperature Vegetation Precipitation Dryness Index(TVPDI), Vegetation Health Index(VHI), and Temperature Condition Index (TCI). An evaluation was conducted to assess the effectiveness of seven drought indicators, such as DSI, ESI, TVPDI, VSWI, etc., in capturing the changes in SM in Mu Us Sandy Land. The research results indicated that DSI and ESI had the highest accuracy, while TVDI and VSWI showed relatively lower accuracy. However, their smaller fluctuations in the time series demonstrated stronger adaptability to different regions. Additionally, the delayed impact of aridity indices on soil moisture, variable attributes, temperature, and vegetation coverage in sandy land and grassland areas with low, medium, and high coverage all contributed to the effectiveness of the four aridity indices (DSI, ESI, VSWI, and TVPDI) in capturing the dynamics of soil moisture. The primary element that affects the effectiveness of TVDI is the divergence of the relationship curve between Land Surface Temperature (LST) and Normalized Difference Vegetation Index (NDVI), which is a kind of deterioration. This paper presents a very efficient approach for monitoring soil moisture dynamics in dry and semi-arid regions. It also analyzes the patterns of soil moisture changes, offering valuable scientific insights for environmental monitoring and ecological enhancement.

The objective of this study is to assess the suitability of the AquaCrop model for growing maize using brackish water irrigation in Northwest China. Additionally, this study aims to examine how maize utilizes water in various soil layers when irrigated with varying water qualities. The AquaCrop model was calibrated and verified using experimental data from the years 2022 and 2023 in this research. (1) The findings indicated that the AquaCrop model effectively simulated the canopy cover, biomass, and yield of maize when irrigated with brackish water. The validation year’s R², MAPE, and RMSE values for canopy cover, biomass, and yield of maize were 0.95, 5.36%, and 4.77%, respectively. For biomass, the R², MAPE, and RMSE values were 0.91, 16.61%, and 2.12 t·hm⁻², respectively. For yield, the R², MAPE, and RMSE values were 0.84, 3.62%, and 0.42 t·hm⁻², respectively. (2) Irrigation with water of high mineral content, measured at 1.6 ds/m, as well as with fresh water over the whole reproductive period, resulted in an increased reliance on groundwater for maize cultivation. There was no notable disparity in the usage of various soil layers between the irrigation with alternating freshwater and brackish water. (3) The AquaCrop model simulated the effects of seven different irrigation water quality treatments. It was shown that using water with mineralization levels of 0.5 and 0.8 ds/m resulted in decreased freshwater use without causing a substantial decrease in maize yield and biomass.