Thriving arid oasis urban agglomerations: Optimizing ecosystem services pattern under future climate change scenarios using dynamic Bayesian network.

The effects of global climate change and human activities are anticipated to significantly impact ecosystem services (ESs), particularly in urban agglomerations of arid regions. This paper proposes a framework integrating the dynamic Bayesian network (DBN), system dynamics (SD) model, patch generation land use simulation (PLUS) model, and the Integrated Valuation of Ecosystem Services and Tradeoffs (InVEST) model for predicting land use change and optimizing ESs spatial patterns that is built upon the SSP-RCP scenarios from CMIP6. This framework is applied to the oasis urban agglomeration on the northern slope of the Tianshan Mountains in Xinjiang (UANSTM), China. The findings indicate that both the SD model and PLUS model can accurately forecast the distribution of future land use. The SD model shows a relative error of less than 2.32%, while the PLUS model demonstrates a Kappa coefficient of 0.89. The land use pattern displays obvious spatial heterogeneity under different climate scenarios. The expansion of cultivated land and construction land is the main form of land use change in UANSTM in the future. The DBN model proficiently simulates the interactive relationships between ESs and diverse factors. The classification error rates for net primary productivity (NPP), habitat quality (HQ), water yield (WY), and soil retention (SR) are 20.04%, 3.47%, 4.45%, and 13.42%, respectively. The prediction and diagnosis of DBN determine the optimal ESs development scenario and the optimal ESs region in the study area. It is found that the majority of ESs in UANSTM are predominantly influenced by natural factors with the exception of HQ. The socio-economic development plays a minor role in such urban agglomerations. This study offers significant insights that can contribute to the fields of ecological protection and land use planning in arid urban agglomerations worldwide.

Convergence and divergence emerging in climatic controls of polynomial trends for northern ecosystem productivity over 2000-2018.

Recent rapid warming has caused uneven impacts on the composition, structure, and functioning of northern ecosystems. It remains unknown how climatic drivers control linear and non-linear trends in ecosystem productivity. Based on a plant phenology index (PPI) product at a spatial resolution of 0.05° over 2000-2018, we used an automated polynomial fitting scheme to detect and characterize trend types (i.e., polynomial trends and no-trends) in the yearly-integrated PPI (PPIINT) for northern (> 30°N) ecosystems and their dependence on climatic drivers and ecosystem types. The averaged slope for the linear trends (p < 0.05) of PPIINT was positive across all the ecosystems, among which deciduous broadleaved forests and evergreen needle-leaved forests (ENF) showed the highest and lowest mean slopes, respectively. More than 50% of the pixels in ENF, arctic and boreal shrublands, and permanent wetlands (PW) had linear trends. A large fraction of PW also showed quadratic and cubic trends. These trend patterns agreed well with estimates of global vegetation productivity based on solar-induced chlorophyll fluorescence. Across all the biomes, PPIINT in pixels with linear trends showed lower mean values and higher partial correlation coefficients with temperature or precipitation than in pixels without linear trends. Overall, our study revealed the emergence of latitudinal convergence and divergence in climatic controls on the linear and non-linear trends of PPIINT, implying that northern shifts of vegetation and climate change may potentially increase the non-linear nature of climatic controls on ecosystem productivity. These results can improve our understanding and prediction of climate-induced changes in plant phenology and productivity and facilitate sustainable management of ecosystems by accounting for their resilience and vulnerability to future climate change.

Assessing spatial connectivity effects on daily streamflow forecasting using Bayesian-based graph neural network.

Data-driven models have been widely developed and achieved impressive results in streamflow prediction. However, the existing data-driven models mostly focus on the selection of input features and the adjustment of model structure, and less on the impact of spatial connectivity on daily streamflow prediction. In this paper, a basin network based on graph-structured data is constructed by considering the spatial connectivity of different stations in the real basin. Furthermore, a novel graph neural network model, variational Bayesian edge-conditioned graph convolution model, which consists of edge-conditioned convolution networks and variational Bayesian inference, is proposed to assess the spatial connectivity effects on daily streamflow forecasting. The proposed graph neural network model is applied to forecast the next-day streamflow of a hydrological station in the Yangtze River Basin, China. Six comparative models and three comparative experimental groups are used to validate model performance. The results show that the proposed model has excellent performance in terms of deterministic prediction accuracy (NSE ≈ 0.980, RMSE≈1362.7 and MAE ≈ 745.8) and probabilistic prediction reliability (ICPC≈0.984 and CRPS≈574.1), which demonstrates that establishing appropriate connectivity and reasonably identifying connection relationships in the basin network can effectively improve the deterministic and probabilistic forecasting performance of the graph convolutional model.

Prediction of Greenhouse Tomato Crop Evapotranspiration Using XGBoost Machine Learning Model.

Crop evapotranspiration estimation is a key parameter for achieving functional irrigation systems. However, ET is difficult to directly measure, so an ideal solution was to develop a simulation model to obtain ET. There are many ways to calculate ET, most of which use models based on the Penman−Monteith equation, but they are often inaccurate when applied to greenhouse crop evapotranspiration. The use of machine learning models to predict ET has gradually increased, but research into their application for greenhouse crops is relatively rare. We used experimental data for three years (2019−2021) to model the effects on ET of eight meteorological factors (net solar radiation (Rn), mean temperature (Ta), minimum temperature (Tamin), maximum temperature (Tamax), relative humidity (RH), minimum relative humidity (RHmin), maximum relative humidity (RHmax), and wind speed (V)) using a greenhouse drip irrigated tomato crop ET prediction model (XGBR-ET) that was based on XGBoost regression (XGBR). The model was compared with seven other common regression models (linear regression (LR), support vector regression (SVR), K neighbors regression (KNR), random forest regression (RFR), AdaBoost regression (ABR), bagging regression (BR), and gradient boosting regression (GBR)). The results showed that Rn, Ta, and Tamax were positively correlated with ET, and that Tamin, RH, RHmin, RHmax, and V were negatively correlated with ET. Rn had the greatest correlation with ET (r = 0.89), and V had the least correlation with ET (r = 0.43). The eight models were ordered, in terms of prediction accuracy, XGBR-ET > GBR-ET > SVR-ET > ABR-ET > BR-ET > LR-ET > KNR-ET > RFR-ET. The statistical indicators mean square error (0.032), root mean square error (0.163), mean absolute error (0.132), mean absolute percentage error (4.47%), and coefficient of determination (0.981) of XGBR-ET showed that XGBR-ET modeled daily ET for greenhouse tomatoes well. The parameters of the XGBR-ET model were ablated to show that the order of importance of meteorological factors on XGBR-ET was Rn > RH > RHmin> Tamax> RHmax> Tamin> Ta> V. Selecting Rn, RH, RHmin, Tamax, and Tamin as model input variables using XGBR ensured the prediction accuracy of the model (mean square error 0.047). This study has value as a reference for the simplification of the calculation of evapotranspiration for drip irrigated greenhouse tomato crops using a novel application of machine learning as a basis for an effective irrigation program.

Assessing landscape fragmentation in a desert-oasis region of Northwest China: patterns, driving forces, and policy implications for future land consolidation.

Landscape fragmentation is considered a serious threat to eco-environmental integrity and socioeconomic development. Although many studies have focused on landscape fragmentation resulting from agricultural production and urbanization, landscape fragmentation from the aspects of patterns, driving forces, and the policy perspective of ecosystems has rarely been investigated. Oases, as a unique landscape, face severe fragmentation in arid and semiarid regions. This study applied a combination of approaches, including remote sensing image interpretations, landscape fragmentation metrics, and community surveys, to analyze patterns and their driving forces, as well as the policy implications for future land consolidation, in the Hotan oasis of Northwest China from the space and time perspectives. Results show that the frequent occurrence of summer flood events changes the patch number, density, size, and splitting degree of oasis-desert ecotone vegetation. The socioeconomic factors including total population and irrigation area are more important driving forces on oasis landscape fragmentation, compared with natural factors such as temperature and precipitation. Rural expansion, road and canal system developments caused by population growth, and the rising number of households increase oasis landscape fragmentation. Rapid economic development, such as agricultural expansion and urbanization, has imposed the intensification of landscape fragmentation. Fragmentation reaches peak when agricultural development makes up 40-50% of study area. Rural residential reconstruction and farmland transfer policies facilitate the intensive utilization of land toward oasis fragmentation solutions, but many factors, such as landholders' household characteristics and living conditions, are partly responsible for the challenges in land consolidation. This study also demonstrates that intense human activities pose a great threat for land consolidation and sustainable development of oasis landscape.

Drivers of the water use efficiency changes in China during 1982-2015.

This study investigates the drivers of water use efficiency (WUE), a key metric of water resources management, and its changes over eight regions across China from 1982 to 2015 based on gross primary production (GPP) and actual evapotranspiration (AET) datasets. The order of seasonal change of WUE from large to small is autumn, summer, spring and winter. The drivers include seven variables, air temperature, specific humidity, precipitation, short-wave radiation, Normalized Difference Vegetation Index (NDVI), soil moisture and CO2. Our analysis suggests that the sensitivity of annual average NDVI to WUE changes was high nationwide, but there were some differences in seasonal scales. The annual average contribution of air temperature and CO2 affecting WUE change was relatively high in China's largest area (SW, SE, E, NP). Other influencing factors were only relatively high in the local area. Seasonally, NDVI is the driving factor with the highest contribution rate in summer and autumn for NC and NW region. The seasonal contribution rates of driving factors in other regions are significantly different. For the study period (1982-2015), the shrubland ecosystem had the highest annual WUE followed by forest and cropland. The WUE of the farmland ecosystem was higher than that of the grassland ecosystem in most areas.

Estimating evapotranspiration by coupling Bayesian model averaging methods with machine learning algorithms.

Evapotranspiration (ET) is one of the most important components of global hydrologic cycle and has significant impacts on energy exchange and climate change. Numerous models have been developed to estimate ET so far; however, great uncertainties in models still require considerations. The aim of this study is to reduce model errors and uncertainties among multi-models to improve daily ET estimate. The Bayesian model averaging (BMA) method is used to assemble eight ET models to produce ET with Landsat 8 satellite data, including four surface energy balance models (i.e., SEBS, SEBAL, SEBI, and SSEB) and four machine learning algorithms (i.e., polymars, random forest, ridge regression, and support vector machine). Performances of each model and BMA method were validated through in situ measurements of semi-arid region. Results indicated that the BMA method outperformed all eight single models. The four most important models obtained by the BMA method were ranked by random forest, SVM, SEBS, and SEBAL. The BMA method coupled with machine learning can significantly improve the accuracy of daily ET estimate, reducing uncertainties among models, and taking different intrinsic benefits of empirically and physically based models to obtain a more reliable ET estimate.

Acceleration of soil salinity accumulation and soil degradation due to greenhouse cultivation: a survey of farmers' practices in China.

Soil environment and water quality face large pressure due to the rapid expansion of greenhouse cultivation in China. However, studies rarely provide the linkage between farmers' practices and soil degradation in greenhouse cultivation field. In this study, a field survey and sampling of greenhouse cultivation soil were conducted in five regions of China to investigate the accumulation and variation characteristics of soil ion compositions in the field. First, the pH, ion compositions, and electrical conductivity (EC) of 132 composite soil samples were analyzed. Second, farmers' practices with regard to fertilizer, crop yield, and soil degradation processes were surveyed. Lastly, soil nutrient status was evaluated by different grades, and the principal component analysis method was used to analyze the main sources of soil ion compositions. Results of the study reveal the following: (1) Enrichment of greenhouse soil nutrient was mainly caused by excessive fertilization, which introduced the secondary salinization phenomenon for 3-5 years in plastic greenhouse and 1-3 years in multispan greenhouse. (2) Significant changes between the EC and salt ion composition of open soil and greenhouse cultivated soil were observed. The contents of nitrate nitrogen and ammonium nitrogen in the greenhouse soil were high. (3) After a certain period of cultivation in the greenhouse, salt accumulation, pH decline, and varying degrees of acidification were observed in the soil profile. The relationship between soil pH and EC values indicated that the balance of soil compositions was broken. The recommended methods for sustaining greenhouse cultivation include balanced fertilization, rotation practices, and reasonable water utilization in the field.

Characteristics of salt contents in soils under greenhouse conditions in China.

Greenhouse cultivation is expanding in China due to high production efficiency and greater economic benefits. Although the accumulation of soil salinity and nutrients has been observed in greenhouse cultivation areas, the linkage between soil salinity, soil major ions, and farm practices is not clear in China. Few studies have examined soil salinity accumulated in soil layers; thus, a broad investigation is needed in order to understand the potential causes of soil salinity in greenhouse soil. In this study, a short review was given to show the salt contents and the major ion under greenhouse conditions in China. Then, we analyzed a total of 132 soil samples from different parts of China in terms of their soil major ions and nutrient components and investigated the relevant farm practices. Based on survey data from three different types of cultivation areas (open farmland, plastic greenhouses, and multispan greenhouses), we found that cultivation in both greenhouse types resulted in a significant increase in salt content and a decrease in soil pH values, a pattern not shown in open farmland. The linkage between soil salinity and cultivation type was confirmed by soil salinity classification. The proportion of each ion in the soil salt differs significantly between the different management methods, but the variation range of the main ions ranged from - 23.3 to 225.6% for multispan greenhouses and - 22.6 to 430.5% for plastic greenhouses. In addition, the salt source in greenhouses is not unique to those methods, suggesting that different growing practices cause the differences in ion concentration. Removing greenhouse covers during the rainy season can avoid further accumulations of salt, but the subsequent rinsing of soil can lead to the deeper salt accumulations. In addition, increasing salt content may lead to decreasing pH once the natural salt balance is altered. These results show that the soil salinization produced by greenhouse cultivation cannot be ignored.

Experimental study on soluble chemical transfer to surface runoff from soil.

Prevention of chemical transfer from soil to surface runoff, under condition of irrigation and subsurface drainage, would improve surface water quality. In this paper, a series of laboratory experiments were conducted to assess the effects of various soil and hydraulic factors on chemical transfer from soil to surface runoff. The factors include maximum depth of ponding water on soil surface, initial volumetric water content of soil, depth of soil with low porosity, type or texture of soil and condition of drainage. In the experiments, two soils, sand and loam, mixed with different quantities of soluble KCl were filled in the sandboxes and prepared under different initial saturated conditions. Simulated rainfall induced surface runoff are operated in the soils, and various ponding water depths on soil surface are simulated. Flow rates and KCl concentration of surface runoff are measured during the experiments. The following conclusions are made from the study results: (1) KCl concentration in surface runoff water would decrease with the increase of the maximum depth of ponding water on soil surface; (2) KCl concentration in surface runoff water would increase with the increase of initial volumetric water content in the soil; (3) smaller depth of soil with less porosity or deeper depth of soil with larger porosity leads to less KCl transfer to surface runoff; (4) the soil with finer texture, such as loam, could keep more fertilizer in soil, which will result in more KCl concentration in surface runoff; and (5) good subsurface drainage condition will increase the infiltration and drainage rates during rainfall event and will decrease KCl concentration in surface runoff. Therefore, it is necessary to reuse drained fertile water effectively during rainfall, without polluting groundwater. These study results should be considered in agriculture management to reduce soluble chemical transfer from soil to surface runoff for reducing non-point sources pollution.

Simplified continuous simulation model for investigating effects of controlled drainage on long-term soil moisture dynamics with a shallow groundwater table.

Accurate modeling of soil water content is required for a reasonable prediction of crop yield and of agrochemical leaching in the field. However, complex mathematical models faced the difficult-to-calibrate parameters and the distinct knowledge between the developers and users. In this study, a deterministic model is presented and is used to investigate the effects of controlled drainage on soil moisture dynamics in a shallow groundwater area. This simplified one-dimensional model is formulated to simulate soil moisture in the field on a daily basis and takes into account only the vertical hydrological processes. A linear assumption is proposed and is used to calculate the capillary rise from the groundwater. The pipe drainage volume is calculated by using a steady-state approximation method and the leakage rate is calculated as a function of soil moisture. The model is successfully calibrated by using field experiment data from four different pipe drainage treatments with several field observations. The model was validated by comparing the simulations with observed soil water content during the experimental seasons. The comparison results demonstrated the robustness and effectiveness of the model in the prediction of average soil moisture values. The input data required to run the model are widely available and can be measured easily in the field. It is observed that controlled drainage results in lower groundwater contribution to the root zone and lower depth of percolation to the groundwater, thus helping in the maintenance of a low level of soil salinity in the root zone.

Changes and Relationships of Climatic and Hydrological Droughts in the Jialing River Basin, China.

The comprehensive assessment of climatic and hydrological droughts in terms of their temporal and spatial evolutions is very important for water resources management and social development in the basin scale. To study the spatial and temporal changes of climatic and hydrological droughts and the relationships between them, the SPEI and SDI are adopted to assess the changes and the correlations of climatic and hydrological droughts by selecting the Jialing River basin, China as the research area. The SPEI and SDI at different time scales are assessed both at the entire Jialing River basin and at the regional levels of the three sub basins. The results show that the SPEI and SDI are very suitable for assessing the changes and relationships of climatic and hydrological droughts in large basins. Based on the assessment, for the Jialing River basin, climatic and hydrological droughts have the increasing tendency during recent several decades, and the increasing trend of climatic droughts is significant or extremely significant in the western and northern basin, while hydrological drought has a less significant increasing trend. Additionally, climatic and hydrological droughts tend to increase in the next few years. The results also show that on short time scales, climatic droughts have one or two months lag impact on hydrological droughts in the north-west area of the basin, and have one month lag impact in south-east area of the basin. The assessment of climatic and hydrological droughts based on the SPEI and SDI could be very useful for water resources management and climate change adaptation at large basin scale.

Global sensitivity analysis for an integrated model for simulation of nitrogen dynamics under the irrigation with treated wastewater.

As the amount of water resources that can be utilized for agricultural production is limited, the reuse of treated wastewater (TWW) for irrigation is a practical solution to alleviate the water crisis in China. The process-based models, which estimate nitrogen dynamics under irrigation, are widely used to investigate the best irrigation and fertilization management practices in developed and developing countries. However, for modeling such a complex system for wastewater reuse, it is critical to conduct a sensitivity analysis to determine numerous input parameters and their interactions that contribute most to the variance of the model output for the development of process-based model. In this study, application of a comprehensive global sensitivity analysis for nitrogen dynamics was reported. The objective was to compare different global sensitivity analysis (GSA) on the key parameters for different model predictions of nitrogen and crop growth modules. The analysis was performed as two steps. Firstly, Morris screening method, which is one of the most commonly used screening method, was carried out to select the top affected parameters; then, a variance-based global sensitivity analysis method (extended Fourier amplitude sensitivity test, EFAST) was used to investigate more thoroughly the effects of selected parameters on model predictions. The results of GSA showed that strong parameter interactions exist in crop nitrogen uptake, nitrogen denitrification, crop yield, and evapotranspiration modules. Among all parameters, one of the soil physical-related parameters named as the van Genuchten air entry parameter showed the largest sensitivity effects on major model predictions. These results verified that more effort should be focused on quantifying soil parameters for more accurate model predictions in nitrogen- and crop-related predictions, and stress the need to better calibrate the model in a global sense. This study demonstrates the advantages of the GSA on a more complete analysis of model input parameters and their interactions on the model output for nitrogen modeling.