The stability of riverbanks in cold regions is affected by the freeze-thaw action. The freeze-thaw process causes changes in the moisture content, friction angle, and cohesiveness of the bank material. Together with the freeze-thaw effect, seepage pressure influenced by the changing water levels, and the bank slope are the key factors affecting bank stability. A limit equilibrium bank stability model considering the infiltration water pressure effect is developed and applied to the Shisifenzi section of Inner Mongolia reach of the Yellow River. Laboratory tests of field samples with moisture contents of 15%, 18%, 21%, and saturated showed that the freeze-thaw action reduced the degree of saturation by 34.37 %, 30.71%, 32.48%, and 46.23%, respectively, accompanied by reductions in the internal friction angles by 1.78%, 2.74%, 6.33%, and 5.32%. These changes resulted in a 24.35% to 29.13% reduction in the safety factor of bank stability. Together with seasonal variations in the water levels the field data showed that the bank stability safety factor in the study site increases gradually through the melting period, dry period, wet period, flooding period, and low flow period. The slope stability safety factor increases with the stage of the river but decreases with the groundwater level.

We aim to study the impact of microalgae fertilizer on soil nutrients, water conservation and crop yield and quality while also determining the optimal ratio of microalgae fertilizer to chemical fertilizer. Using “Xinoufen No.9” tomatoes as the test subject, we conducted pot experiments with four different treatments: control with 100% chemical fertilizer (CK), T1 (25% microalgae fertilizer + 75% regular chemical fertilizer), T2 (75% microalgae fertilizer + 25% regular chemical fertilizer) and T3 (100% microalgae fertilizer). The results show that an increased application of microalgae fertilizer enhanced the soil organic matter, ammonium nitrogen, available phosphorus and potassium content. T3 showed the most improvement followed by T2. The co-application of microalgae fertilizer with chemical fertilizer can significantly increase the stem girth, plant height and yield of tomatoes. At the same time, microalgae fertilizer effectively regulates leaf stomatal conductance, promoting tomato leaf respiration. As the stomatal conductance increases, the transpiration rate and net photosynthesis rate of all treatments improve, followed by a decline in intercellular CO₂ concentration, with T2 exhibiting the best performance. Among all treatments, T2 treatment yielded the highest per-plant production (0.630 kg), followed by T3 (0.521 kg). This is because the microalgae fertilizer promotes the distribution of photosynthetic products to the fruit, enhancing the yield and quality of tomatoes. Additionally, the microalgae fertilizer also increases the content of soluble sugars, soluble protein, vitamin C and lycopene in the fruit while reducing the nitrate content. Compared to the control group CK, T2 increases the content of soluble sugars, vitamins and lycopene by 26.74%, 39.29% and 158.31%, respectively. Microalgae fertilizer also helps to improve soil water and thermal conditions, enhancing the water-use efficiency of tomatoes. Compared to CK, the water-use efficiency of T2 treatment increased by 54.05%. Correlation analysis indicates that water and fertilizer factors significantly affect tomato yield, with a correlation exceeding 70%. The net photosynthesis and transpiration rates significantly influence fruit quality, with correlations above 80%. By applying microalgae fertilizer, the efficiency of water and fertilizer use can be effectively improved, thus achieving the goal of water conservation and quality enhancement. Therefore, through comprehensive analysis, using the membership function method of indicators such as soil environment, crop yield, fruit quality and water-use efficiency, it is concluded that T2 is the optimal fertilization treatment. This study provides theoretical support for the application of microalgae biofertilizer technology in the cultivation of tomatoes and other vegetables in the northern, cold and arid regions.

Concern has been expressed regarding the impacts of climate change on river ice and ice jam formation in cold regions. Ice jams are easily initiated in bends and narrow channels and cause disasters. In this study, observations and remote sensing monitoring are used to study the freeze-up ice jam formation of bends. Sediment transport and freezing process of the river interact, influencing bed changes profile and sedimentary budget. River ice processes, channel evolution, ice hydro-thermodynamics, and ice jam accumulation are explored. The results show that the channel topography determines the river thalweg, and that the channel elevation interacts with the river ice through sediment transport. The channel shrinkage increases the probability of ice jam, and the sharp bend is prone to ice jam formation. Under the effect of secondary circulation flow in the bend and in the outer bank, the juxtaposed freeze-up and the hummocky ice cover occur in the same location, and frazil ice accumulates under the junction of the main channel and the shoals. Affected by the increase of the hydraulic slope and the velocity downstream, open water reaches develops downstream of the ice accumulation. An open water section is emerged upstream of the bend, due to the ice deposition, and partly cut-off supply of the frazil.

Net ecosystem productivity (NEP) is a critical indicator for characterizing the carbon cycle dynamics within terrestrial ecosystems. This study employs six different combinations of methods for calculating Net Primary Productivity (NPP) and heterotrophic soil respiration Rh) to estimate monthly NEP values in Inner Mongolia from 2001 to 2021. The carbon flux observation data obtained through the eddy covariance method are used to validate and evaluate these combinations, and the best NEP estimation model combination is selected, and the spatiotemporal distribution patterns of NEP along with its primary driving factors are analyzed. Results show that: 1) The NEP estimates derived from MODIS NPP combined with the Global Soil Respiration Model (GSMSR) and Bond-Lamberty’s Rs-Rh relationship model exhibit a strong correlation with validated data; 2) The NEP in Inner Mongolia shows a significant increasing trend, with an annual average value of 168.73 gC·m⁻²·a⁻¹, or 177.57 gC·m⁻²·a⁻¹ when excluding barren. Forests, croplands, and grasslands are identified as the primary carbon sinks during the growing season, with average NEP values of 84.81, 46.41, and 32.95 gC·m⁻²·mth⁻¹, respectively; 3) Precipitation is the dominant meteorological factor driving the spatiotemporal variations of NEP across the region, contributing 72.29% to NEP during the growing season. Additionally, over 80% of areas influenced by human activities exhibit a positive impact on NEP; 4) The interannual and growing season increases in NEP are primarily attributed to climate change and anthropogenic activities, which account for 57% and 66.3% of NEP variations, respectively. These effects are particularly pronounced in the eastern forested regions and central grasslands of Inner Mongolia. The findings of this study provide valuable insights for regional carbon sink management and ecological environment protection.

Extraction of lake water bodies from remote sensing images provides data support and a scientific basis for water resource management, environmental monitoring and protection, climate change research, disaster monitoring and risk assessment, urban planning, and land use. Aiming at the problems of insufficient spatial details, edge recognition, and poor anti-noise performance in the existing lake water body extraction models, a lake water body extraction model AMFF-LWBENet (A Lake Water Body Extraction Network Based on ASPP and Multi-scale Feature Fusion) is proposed. The model uses an encoder-decoder structure based on techniques such as deep separable convolution and ASPP, and uses ResNet50 deep residual network and multiscale dense fusion (MDF) module for better extraction and fusion of multiscale features. After that, the positional correlation and channel correlation in the feature map are paid attention to by DANet, which can reduce the interference of noise information on the lake boundary and improve the accuracy of the network in segmenting the lake boundary. Finally, bilinear interpolation is utilized to up-sample the feature map and feature fusion through the cross-layer feature fusion (CFF) module, which finally realizes the accurate extraction of the lake water body. Comparing with other state-of-the-art semantic segmentation network models on both self-built dataset (W-H) and public dataset (TP), AMFF-LWBENet achieved the highest model accuracy. The MIoU values of the model on the two datasets are 97.52% and 97.5%, respectively. The experimental results show that AMFF-LWBENet is an excellent network for lake water body extraction.