Spatiotemporal assessment of groundwater quality and quantity using geostatistical and ensemble artificial intelligence tools.

The study investigated the spatiotemporal relationship between surface hydrological variables and groundwater quality/quantity using geostatistical and AI tools. AI models were developed to estimate groundwater quality from ground-based measurements and remote sensing images, reducing reliance on laboratory testing. Different Kriging techniques were employed to map ground-based measurements and fill data gaps. The methodology was applied to analyze the Maragheh aquifer in northwest Iran, revealing declining groundwater quality due to industrial. discharges and over-extraction. Spatiotemporal analysis indicated a relationship between groundwater depth/quality, precipitation, and temperature. The Root Mean Square Scaled Error (RMSSE) values for all variables ranged from 0.8508 to 1.1688, indicating acceptable performance of the semivariogram models in predicting the variables. Three AI models, namely Feed-Forward Neural Networks (FFNNs), Support Vector Regression (SVR), and Adaptive Neural Fuzzy Inference System (ANFIS), predicted groundwater quality for wet (June) and dry (October) months using input variables such as groundwater depth, temperature, precipitation, Normalized Difference Vegetation Index (NDVI), and Digital Elevation Model (DEM), with Groundwater Quality Index (GWQI) as the target variable. Ensemble methods were employed to combine the outputs of these models, enhancing performance. Results showed strong predictive capabilities, with coefficient of determination values of 0.88 and 0.84 for wet and dry seasons. Ensemble models improved performance by up to 6% and 12% for wet and dry seasons, respectively, potentially advancing groundwater quality modeling in the future.

What drives the decrease of glacier surface albedo in High Mountain Asia in the past two decades?

Glacier surface albedo is an important factor affecting glacier ablation, and a positive feedback mechanism has been observed between the surface albedo and mass balance of glaciers. It is important to understand the driving factors and mechanisms of glacier albedo changes (GAC). Based on the MODIS (Moderate resolution imaging spectroradiometer)-derived MOD10A1 and MYD10A1 albedo products, the glacier albedo trends in each MODIS grid cell during each melt season in High-Mountains Asia (HMA) from 2000 to 2020 were calculated. Decreasing glacier albedo trends were found, with a decline rate of 0.25 × 10-2 yr-1; in addition, the GACs exhibited great spatial differences among the 15 subregions. The geographical detector model (GDM) is a new spatial statistical method that can quantitatively reveal the driving forces of climate factors and light-absorbing particles on GAC under single-factor and two-factor interactions. These driving forces can be measured by the corresponding q value. The results showed that on the whole, solid precipitation (snowfall) had the strongest impact on GAC, followed by the glacier surface temperature. The q values of black carbon (BC) and dust were <0.1, but BC or dust had the greatest q value in the 9 subregions. The effects of each factor differed among different elevation zones. The interaction detector indicated that the q value under the influence of two factors was greater than that under a single factor, and the strongest interaction was between snowfall and BC, followed by between snowfall and dust. In 15 subregions, most of the greatest q values in each region corresponded to an interaction with BC or dust. Here, we obtained the main driving factors of GAC in different regions and emphasized the interactions between climatic factors and light-absorbing particles; these results provide references for further studies of glacier mass balance and surface albedo.

[Variation of irradiance in the arctic pole during the summer].

The variation of irradiance affect the melting rate of the sea ice in the arctic pole, and the research on it is an important component of the global climate change research. The present research was based on the spectrum data collected during the 4th scientific research on the arctic of China in 2010, analyzed the variation of irradiance in the arctic pole during the summer and discussed the reasons for the change. This research shows that many factors lead to the change, among which the weather and the solar elevation angle affect the irradiance directly. The weather factors determine the amount of solar radiation that reached the ground after the absorption and attenuation of the clouds; In high-latitude areas, there is a low solar elevation angle and the attenuation of solar radiation was obvious. Our research shows that the spectrum at shorter wavelength is more sensitive to the changes in altitude of the sun, while the impact of weather on the irradiance increases with wavelength. Moreover, moisture content in the atmosphere also affects the solar radiation reaching the ground and the its impact is in a particular band but not for the entire spectrum range.

[Spectral features analysis of sea ice in the Arctic Ocean].

Sea ice in the Arctic Ocean plays an important role in the global climate change, and its quick change and impact are the scientists' focus all over the world. The spectra of different kinds of sea ice were measured with portable ASD FieldSpec 3 spectrometer during the long-term ice station of the 4th Chinese national Arctic Expedition in 2010, and the spectral features were analyzed systematically. The results indicated that the reflectance of sea ice covered by snow is the highest one, naked sea ice the second, and melted sea ice the lowest. Peak and valley characteristics of spectrum curves of sea ice covered by thick snow, thin snow, wet snow and snow crystal are very significant, and the reflectance basically decreases with the wavelength increasing. The rules of reflectance change with wavelength of natural sea ice, white ice and blue ice are basically same, the reflectance of them is medium, and that of grey ice is far lower than natural sea ice, white ice and blue ice. It is very significant for scientific research to analyze the spectral features of sea ice in the Arctic Ocean and to implement the quantitative remote sensing of sea ice, and to further analyze its response to the global warming.

What caused the spatial heterogeneity of lake ice phenology changes on the Tibetan Plateau?

As the Third Pole of the Earth, the Tibetan Plateau has numerous lakes with seasonal ice cover. The ice phenology of these lakes has undergone remarkable changes in recent years. We obtained the ice phenology records for 71 lakes for the period of 2001 to 2020 and found overall later trends for both freeze-up and break-up dates. As a result, the changes in ice cover duration showed great spatial heterogeneity. Therefore, we analyzed the causes of lake ice phenology changes from two aspects: climate change and lake properties. The results showed that the changes in air temperature dominated the variations in ice phenology, followed by solar radiation. The weakened wind power in the northeastern part of the plateau was favorable for the delay of break-up end dates and the extension of ice cover durations. Furthermore, by changing the lake size and salinity, water balance changes led to aggravated ice phenology changes for some lakes, while for some other lakes, they moderated or even reversed the changes caused by other climatic factors. In general, the spatial inconsistency of changes in multiple climatic factors (especially differences between the northeastern and southwestern parts) during the 20 years was the main reason for the heterogeneity of lake ice phenology changes on the Tibetan Plateau. This study preliminarily summarized some of the effects of climate change and lake properties on lake ice phenology, and the results are important for understanding the physical mechanism of lake ice phenology changes under climate change.

Monitoring ice variations in Qinghai Lake from 1979 to 2016 using passive microwave remote sensing data.

Lake ice is a sensitive indicator of climate change. Based on the disparities between the brightness temperatures of lake ice and water, passive microwave data can be used to monitor the ice variations of a lake. With focus on the analysis of long time series variability of lake ice, this study extracts four characteristic dates related to lake ice (the annual freeze start, freeze completion, ablation start and ablation completion dates) for Qinghai Lake from 1979 to 2016 using Scanning Multichannel Microwave Radiometer (SMMR) and Special Sensor Microwave Imager (SSM/I) passive microwave brightness temperature data. The corresponding freezing duration, ablation duration, complete freezing duration and ice coverage duration are calculated. Applying Moderate Resolution Imaging Spectroradiometer (MODIS) daily snow products, the accuracy of the results derived from passive microwave data is validated. The validation analysis shows a strong agreement (R2 ranges from 0.70 to 0.85, mean absolute error (MAE) ranges from 2.25 to 3.94days) in the freeze start, ablation start, and ablation completion dates derived from the MODIS data and passive microwave data; the ice coverage duration also has a small error (relative error (RE)=2.95%, MAE=3.13days), suggesting that the results obtained from passive microwave data are reliable. The results show that the freezing dates of Qinghai Lake have been delayed and the ablation dates have advanced. Over 38years, the freeze start date and freeze completion date have been pushed back by 6.16days and 2.27days, respectively, while the ablation start date and ablation completion date have advanced by 11.24days and 14.09days, respectively. The freezing duration and ablation duration have shortened by 3.89days and 2.85days, respectively, and the complete freezing duration and ice coverage duration have shortened by 14.84days and 21.21days, respectively. There is a significant negative correlation between the ice coverage duration and the mean air temperature in winter.