Assessing how land use change will affect water production ecosystem services is essential to developing sound water resource management and ecosystem conservation. The results of a coordination analysis of land-use intensity and water yield based on future land-use simulation projections are useful for future land-use planning. To effectively assess water production rates, the PLUS and InVEST models were used to dynamically assess the changes in water production occurring in the urban agglomeration on the northern slopes of the Tianshan Mountains from 2000 to 2030 under different scenarios of land-use change. The results show that the water-production rates in the study area from 2000 to 2020 were 517.26 × 10⁶ m³, 582.28 × 10⁶ m³, and 456 × 10⁶ m³, showing an increasing and then decreasing trend, with the water production function decreasing from the foothills of the Tianshan Mountains to the north and south, with values of 509.10 × 10⁶ m³³, 510.90 × 10⁶ m³, and 502.28 × 10⁶ m³ being presented for the three scenarios in 2030. The rapid economic development scenario presents the lowest water yield values and the ecological conservation development scenario presents the highest water yield values. Changes in water production rates are closely related to changes in land use, which can be verified further by an analysis of the coordination between land- use intensity and water production. For this study area, the ecological conservation development scenario may be more in line with the future urban development pattern, and the results of the present study can provide some scientific references for land-use planning.

Ecosystems in arid areas are under pressure from human activities and the natural environment. Long-term monitoring and evaluation of arid ecosystems are essential for achieving the goal of sustainable development. The Xinjiang Uygur Autonomous Region (Xinjiang) is a typical arid region located in Northwest China with a relatively sensitive ecosystem. Under the support of the Google Earth Engine (GEE) cloud platform’s massive data collection, the remote sensing ecological index (RSEI) from 2000 to 2020, both in summer and spring, is established, and the variation trend of the ecological quality in Xinjiang is evaluated by coefficient of variation (CV), Sen’s slope analysis, Mann–Kendall trend test (M–K test) and Hurst index. In addition, a partial correlation analysis is processed between RSEI and selected climatic factors, including precipitation and temperature, to find out the mode of correlation between ecological quality and the natural climate. In the last two decades the following has become apparent: (1) The RSEI values of Xinjiang have been relatively low and unstable both in summer and spring, with a trend toward increasing; (2) The distribution characteristics of RSEI levels both in summer and spring have been similar; low levels were concentrated in the desert and wilderness, while high levels were concentrated around the oasis; (3) The ecological quality in Xinjiang has been relatively stable, with a trend of sustained increase both in summer and spring. There was also a small area of sustained decrease around the Junggar Basin and Turpan Basin in summer and a small area of significant decrease in the center of the Taklamakan Desert in spring; (4) In summer, the precipitation has obviously positively correlated in the Southwest. The temperature has obviously positively correlated in the northwestern part; in spring, the precipitation has obviously positively correlated in the Western part; the temperature has obviously positively correlated in the oasis around the Yili River Basin and Tarim Basin.

It has become undeniable that global land surface temperature (LST) has continued to rise in recent years. The threat of extreme heat to humans has become self-evident, especially in arid regions. Many studies have clarified the temperature rise/fall mechanism of LST from the perspective of influencing factors. However, there are few studies on mitigating LST from the standpoint of regional networks. This paper first combines Landsat 8 with Sentinel-2 imagery for LST downscaling based on the Google Earth engine as a way to match local climate zone (LCZ) with 17 classification types. Then, the thermal environment resistance surface is constructed according to LCZ, and the essential cold sources are identified using morphological spatial pattern analysis (MSPA) and circuit theory to form the thermal environment green corridor and obtain the pinch point and barrier point areas. The results show that (1) The downscaling of LST based on random forest (RF) for the Urumqi–Changji–Wujiaqu metropolitan area has an R² of 0.860 and an RMSE of 3.23, with high downscaling accuracy. (2) High temperature (HT), medium temperature (MT), and low temperature (LT) have the largest proportions in the study area; HT dominates in Urumqi, LT in Changji, and MT in Wujiaqu. (3) The natural types (LCZ-D, LCZ-C, and LCZ-F) in the LCZ classification occupy a large area, and the building types are mainly concentrated in Urumqi; LCZ-D, LCZ-G, and LCZ-A contribute the most to the cooling of LST, and LCZ-F, LCZ-C, and LCZ-10 contribute the most to the warming of LST. (4) After identifying critical cold source patches according to MSPA to arrive at 253 green corridors, subsensitive corridors and sensitive corridors need to take certain measures to prevent corridor blockage; pinch point areas, as well as barrier point areas, need to be protected and repaired according to their respective characteristics. In summary, corresponding cooling measures to specific areas can improve the connectivity between cooling sources and slow down the temperature increase of the whole area. This study and experimental approach can provide new insights for urban planners and climate researchers.

An in-depth study of the influence mechanism of oasis land surface temperature (LST) in arid regions is essential to promote the stable development of the ecological environment in dry areas. Based on MODIS, MYD11A2 long time series data from 2003 to 2020, the Mann–Kendall nonparametric test, the Sen slope, combined with the Hurst index, were used to analyze and predict the trend of LST changes in the urban agglomeration on the northern slopes of the Tianshan Mountains. This paper selected nine influencing factors of the slope, aspect, air temperature, normalized vegetation index (NDVI), precipitation (P), nighttime light index (NTL), patch density (PD), mean patch area (AREA_MN), and aggregation index (AI) to analyze the spatial heterogeneity of LST from global and local perspectives using the geodetector (GD) model and multi-scale geo-weighted regression (MGWR) model. The results showed that the average LSTs of the urban agglomeration on the northern slopes of the Tianshan Mountains in spring, summer, autumn, and winter were 31.53 °C, 47.29 °C, 22.38 °C, and −5.20 °C in the four seasons from 2003 to 2020, respectively. Except for autumn, the LST of all seasons showed an increasing trend, bare soil and grass land had a warming effect, and agricultural land had a cooling effect. The results of factor detection showed that air temperature, P, and NDVI were the dominant factors affecting the spatial variation of LST. The interaction detection results showed that the interaction between air temperature and NDVI was the most significant, and the two-factor interaction was more effective than the single-factor effect on LST. The MGWR model results showed that the effects of PD on LST were positively correlated, and the impact of AREA_MN and AI on LST were negatively correlated, indicating that the dense landscape of patches has a cooling effect on LST. Overall, this study provides information for managers to carry out more targeted ecological stability regulations in arid zone oases and facilitates the development of regulatory measures to maintain the cold island effect and improve the environment.