Enhancing Deep Learning Soil Moisture Forecasting Models by Integrating Physics-based Models

Accurate soil moisture(SM)prediction is critical for understanding hydrological processes.Physics-based(PB)models exhibit large uncertainties in SM predictions arising from uncertain parameterizations and insufficient representation of land-surface processes.In addition to PB models,deep learning(DL)models have been widely used in SM predictions recently.However,few pure DL models have notably high success rates due to lacking physical information.Thus,we developed hybrid models to effectively integrate the outputs of PB models into DL models to improve SM predictions.To this end,we first developed a hybrid model based on the attention mechanism to take advantage of PB models at each forecast time scale(attention model).We further built an ensemble model that combined the advantages of different hybrid schemes(ensemble model).We utilized SM forecasts from the Global Forecast System to enhance the convolutional long short-term memory(ConvLSTM)model for 1–16 days of SM predictions.The performances of the proposed hybrid models were investigated and compared with two existing hybrid models.The results showed that the attention model could leverage benefits of PB models and achieved the best predictability of drought events among the different hybrid models.Moreover,the ensemble model performed best among all hybrid models at all forecast time scales and different soil conditions.It is highlighted that the ensemble model outperformed the pure DL model over 79.5%of in situ stations for 16-day predictions.These findings suggest that our proposed hybrid models can adequately exploit the benefits of PB model outputs to aid DL models in making SM predictions.

Time-lagged Effects of the Spring Atmospheric Heat Source over the Tibetan Plateau on Summer Precipitation in Northeast China during 1961–2020:Role of Soil Moisture

The spring atmospheric heat source(AHS)over the Tibetan Plateau(TP)has been suggested to affect the Asian summer monsoon and summer precipitation over South China.However,its influence on the summer precipitation in Northeast China(NEC)remains unknown.The connection between spring TP AHS and subsequent summer precipitation over NEC from 1961 to 2020 is analyzed in this study.Results illustrate that stronger spring TP AHS can enhance subsequent summer NEC precipitation,and higher soil moisture in the Yellow River Valley-North China region(YRVNC)acts as a bridge.During spring,the strong TP AHS could strengthen the transportation of water vapor to East China and lead to excessive rainfall in the YRVNC.Thus,soil moisture increases,which regulates local thermal conditions by decreasing local surface skin temperature and sensible heat.Owing to the memory of soil moisture,the lower spring sensible heat over the YRVNC can last until mid-summer,decrease the land–sea thermal contrast,and weaken the southerly winds over the East Asia–western Pacific region and convective activities over the South China Sea and tropical western Pacific.This modulates the East Asia–Pacific teleconnection pattern,which leads to a cyclonic anomaly and excessive summer precipitation over NEC.

Synergistic Impacts of Indian Ocean SST and Indo-China Peninsula Soil Moisture on the 2020 Record-breaking Mei-yu

The Yangtze River basin(YRB)experienced a record-breaking mei-yu season in June‒July 2020.This unique long-lasting extreme event and its origin have attracted considerable attention.Previous studies have suggested that the Indian Ocean(IO)SST forcing and soil moisture anomaly over the Indochina Peninsula(ICP)were responsible for this unexpected event.However,the relative contributions of IO SST and ICP soil moisture to the 2020 mei-yu rainfall event,especially their linkage with atmospheric circulation changes,remain unclear.By using observations and numerical simulations,this study examines the synergistic impacts of IO SST and ICP soil moisture on the extreme mei-yu in 2020.Results show that the prolonged dry soil moisture led to a warmer surface over the ICP in May under strong IO SST backgrounds.The intensification of the warm condition further magnified the land thermal effects,which in turn facilitated the westward extension of the western North Pacific subtropical high(WNPSH)in June‒July.The intensified WNPSH amplified the water vapor convergence and ascending motion over the YRB,thereby contributing to the 2020 mei-yu.In contrast,the land thermal anomalies diminish during normal IO SST backgrounds due to the limited persistence of soil moisture.The roles of IO SST and ICP soil moisture are verified and quantified using the Community Earth System Model.Their synergistic impacts yield a notable 32%increase in YRB precipitation.Our findings provide evidence for the combined influences of IO SST forcing and ICP soil moisture variability on the occurrence of the 2020 super mei-yu.

Antecedent Precipitation Index to Estimate Soil Moisture and Correlate as a Triggering Process in the Occurrence of Landslides

Landslides are highly dangerous phenomena that occur in different parts of the world and pose significant threats to human populations. Intense rainfall events are the main triggering process for landslides in urbanized slope regions, especially those considered high-risk areas. Various other factors contribute to the process;thus, it is essential to analyze the causes of such incidents in all possible ways. Soil moisture plays a critical role in the Earth’s surface-atmosphere interaction systems;hence, measurements and their estimations are crucial for understanding all processes involved in the water balance, especially those related to landslides. Soil moisture can be estimated from in-situ measurements using different sensors and techniques, satellite remote sensing, hydrological modeling, and indicators to index moisture conditions. Antecedent soil moisture can significantly impact runoff for the same rainfall event in a watershed. The Antecedent Precipitation Index (API) or “retained rainfall,” along with the antecedent moisture condition from the Natural Resources Conservation Service, is generally applied to estimate runoff in watersheds where data is limited or unavailable. This work aims to explore API in estimating soil moisture and establish thresholds based on landslide occurrences. The estimated soil moisture will be compared and calibrated using measurements obtained through multisensor capacitance probes installed in a high-risk area located in the mountainous region of Campos do Jordão municipality, São Paulo, Brazil. The API used in the calculation has been modified, where the recession coefficient depends on air temperature variability as well as the climatological mean temperature, which can be considered as losses in the water balance due to evapotranspiration. Once the API is calibrated, it will be used to extrapolate to the entire watershed and consequently estimate soil moisture. By utilizing recorded mass movements and comparing them with API and soil moisture, it will be possible to determine thresholds, thus enabling anticipation of landslide occurrences.

Response of soil nutrients to terracing and environmental factors in the Loess Plateau of China

Terracing is a widely adopted agricultural practice in mountainous regions around the world that aims to conserve soil and water resources.Soil nutrients play a crucial role in determining soil quality,particularly in landscapes prone to drought.They are influenced by factors such as land-use type,slope aspect,and altitude.In this study,we sought to examine the impact of terracing on soil nutrients(soil organic content(SOC),total nitrogen(TN),nitrate-nitrogen(NO_(3)^(-)-N),ammonium nitrogen(NH_(4)^(+)-N),total phosphorus(TP),available phosphorus(AP),total potassium(TK),and available potassium(AK))and how they vary with environmental factors in the Chinese Loess Plateau.During the growing season,we collected 540 soil samples from the 0 to 100 cm soil layer across five major land-use types,different slope aspects,and varying altitudes.Additionally,a meta-analysis of literature data further corroborated the effective accumulation of soil nutrients through terracing in the Loess Plateau.Our findings are as follows:(1)Terraced fields,regardless of land-use type,showed a significant improvement in SOC and TN content.(2)Soil nutrient contents within terraced fields were predominantly higher on sunny slopes.(3)Terraces at lower altitudes are characterized by elevated SOC concentrations.(4)A meta-analysis of literature data pertaining to terracing and soil nutrients in this region confirmed the effective accumulation of soil nutri-ents through terracing.The elucidated outcomes of this study offer a profound theoretical underpinning for the accurate planning and management of terraces,the scientific utilization of land resources,and the enhancement of land productivity.

A Cloud Framework for High Spatial Resolution Soil Moisture Mapping from Radar and Optical Satellite Imageries

Soil moisture plays an important role in crop yield estimation,irrigation management,etc.Remote sensing technology has potential for large-scale and high spatial soil moisture mapping.However,offline remote sensing data processing is time-consuming and resource-intensive,and significantly hampers the efficiency and timeliness of soil moisture mapping.Due to the high-speed computing capabilities of remote sensing cloud platforms,a High Spatial Resolution Soil Moisture Estimation Framework(HSRSMEF)based on the Google Earth Engine(GEE)platform was developed in this study.The functions of the HSRSMEF include research area and input datasets customization,radar speckle noise filtering,optical-radar image spatio-temporal matching,soil moisture retrieving,soil moisture visualization and exporting.This paper tested the performance of HSRSMEF by combining Sentinel-1,Sentinel-2 images and insitu soil moisture data in the central farmland area of Jilin Province,China.Reconstructed Normalized Difference Vegetation Index(NDVI)based on the Savitzky-Golay algorithm conforms to the crop growth cycle,and its correlation with the original NDVI is about 0.99(P<0.001).The soil moisture accuracy of the random forest model(R 2=0.942,RMSE=0.013 m3/m3)is better than that of the water cloud model(R 2=0.334,RMSE=0.091 m3/m3).HSRSMEF transfers time-consuming offline operations to cloud computing platforms,achieving rapid and simplified high spatial resolution soil moisture mapping.

Study on the Soil Moisture Stress Level in Regulated Deficit Irrigation Experiment

On the basis of discussing the influencing mode of plant moisture stress on plant physiological process and the division of soil moisture availability range, the water suction values partitioning soil moisture were put forward, and then the corresponding water moistures under water stress were obtained by conversing together with characteristic curve of water moisture.

Enhanced Seasonal Predictability of Spring Soil Moisture over the Indo-China Peninsula for Eastern China Summer Precipitation under Non-ENSO Conditions

Seasonal prediction of summer precipitation over eastern China is closely linked to the East Asian monsoon circulation,which is largely affected by the El Niño-Southern Oscillation(ENSO).In this study,results show that spring soil moisture(SM)over the Indo-China peninsula(ICP)could be a reliable seasonal predictor for eastern China summer precipitation under non-ENSO conditions.When springtime SM anomalies are present over the ICP,they trigger a structured response in summertime precipitation over most of eastern China.The resultant south-to-north,tri-polar configuration of precipitation anomalies has a tendency to yield increased(decreased)precipitation in the Yangtze River basin and decreased(increased)in South and North China with a drier(wetter)spring soil condition in the ICP.The analyses show that ENSO exerts a powerful control on the East Asian circulation system in the ENSO-decaying summer.In the case of ENSO forcing,the seasonal predictability of the ICP spring SM for eastern China summer precipitation is suppressed.However,in the absence of the influence of ENSO sea surface temperature anomalies from the preceding winter,the SM anomalies over the ICP induce abnormal local heating and a consequent geopotential height response owing to its sustained control on local temperature,which could,in turn,lead to abnormal eastern China summer precipitation by affecting the East Asian summer monsoon circulation.The present findings provide a better understanding of the complexity of summer climate predictability over eastern China,which is of potential significance for improving the livelihood of the people.

Enhanced soil moisture improves vegetation growth in an arid grassland of Inner Mongolia Autonomous Region, China

Climate change impacts on grasslands that cover a quarter of the global land area, have become unprecedented during the 21~(st) century. One of the important ecological realms, arid grasslands of northern China, which occupy more than 70% of the region's land area. However, the impact of climate change on vegetation growth in these arid grasslands is not consistent and lacks corresponding quantitative research. In this study, NDVI(normalized difference vegetation index) and climate factors including temperature, precipitation, solar radiation, soil moisture, and meteorological drought were analyzed to explore the determinants of changes in grassland greenness in Inner Mongolia Autonomous Region(northern China) during 1982–2016. The results showed that grasslands in Inner Mongolia witnessed an obvious trend of seasonal greening during the study period. Two prominent climatic factors,precipitation and soil moisture accounted for approximately 33% and 27% of grassland NDVI trends in the region based on multiple linear regression and boosted regression tree methods. This finding highlights the impact of water constraints to vegetation growth in Inner Mongolia's grasslands. The dominant role of precipitation in regulating grassland NDVI trends in Inner Mongolia significantly weakened from 1982 to 1996, and the role of soil moisture strengthened after 1996. Our findings emphasize the enhanced importance of soil moisture in driving vegetation growth in arid grasslands of Inner Mongolia, which should be thoroughly investigated in the future.

Enhancing Surface Soil Moisture Estimation through Integration of Artificial Neural Networks Machine Learning and Fusion of Meteorological, Sentinel-1A and Sentinel-2A Satellite Data

For many environmental and agricultural applications, an accurate estimation of surface soil moisture is essential. This study sought to determine whether combining Sentinel-1A, Sentinel-2A, and meteorological data with artificial neural networks (ANN) could improve soil moisture estimation in various land cover types. To train and evaluate the model’s performance, we used field data (provided by La Tuscia University) on the study area collected during time periods between October 2022, and December 2022. Surface soil moisture was measured at 29 locations. The performance of the model was trained, validated, and tested using input features in a 60:10:30 ratio, using the feed-forward ANN model. It was found that the ANN model exhibited high precision in predicting soil moisture. The model achieved a coefficient of determination (R2) of 0.71 and correlation coefficient (R) of 0.84. Furthermore, the incorporation of Random Forest (RF) algorithms for soil moisture prediction resulted in an improved R2 of 0.89. The unique combination of active microwave, meteorological data and multispectral data provides an opportunity to exploit the complementary nature of the datasets. Through preprocessing, fusion, and ANN modeling, this research contributes to advancing soil moisture estimation techniques and providing valuable insights for water resource management and agricultural planning in the study area.

Research Progress of Evaluation Index of Soil Moisture

In modern agriculture, accurate and effective measurements of soil water content lays foundation for promotion on precision irrigation technology and improvement on water use efficiency. The research reviewed soil moisture indices at home and abroad and classified the indices into two categories in order to make prediction on soil moisture and take measures. Specifically, single indices included precipitation index, soil moisture index, and crop drought index and overall indices included supply/demand water index of crops, overall water index, PDSI, crop water shortage index. Soil moisture index was analyzed in terms of advantages and disadvantages,as well as adaptability in agriculture, providing references for relieving and predicting adverse effects on agriculture and formulating scientific policies.

An evaluation of soil moisture from AMSR-E over source area of the Yellow River, China

In this study,in-situ soil moisture measurements are used to evaluate the accuracy of three AMSR-E soil moisture prod ucts from NASA(National Aeronautics and Space Administration),JAXA(Japanese Aerospace Exploration Agency)and VUA(Vrije University Amsterdam and NASA)over Maqu County,Source Area of the Yellow River(SAYR),China.Re sults show that the VUA soil moisture product performs the best among the three AMSR-E soil moisture products in the study area,with a minimum RMSE(root mean square error)of 0.08(0.10)m3/m3 and smallest absolute error of 0.07(0.08)m3/m3 at the grassland area with ascending(descending)data.Therefore,the VUA soil moisture product is used to describe the spatial variation of soil moisture during the 2010 growing season over SAYR.The VUA soil moisture product shows that soil moisture presents a declining trend from east south(0.42 m3/m3)to west north(0.23 m3/m3),with good agreement with a general precipitation distribution.The center of SAYR presents extreme wetness(0.60 m3/m3)dur ing the whole study period,especially in July,while the head of SAYR presents a high level soil moisture(0.23 m3/m3)in July,August and September.

Influence of Vegetation Coverage on Surface Runoff and Soil Moisture in Rainy Season in Dry-hot Valley

[Objective]The research aimed to study the effects of vegetation coverage on the changes of soil moisture in rainy season in dry-hot valley.[Method]The surface runoff and soil moisture of slope with vegetation coverage and bare land in rainy reason in Jinsha River at Yuanmou County of Yunnan Province were observed continuously.Moreover,the statistical analysis was made based on the observation data.[Result]The vegetation coverage could decrease surface runoff and the surface runoff on bare land（CK） was 22 times as the plot with vegetation coverage.The soil water content in 0-180 cm layer with vegetation coverage increased by 37.8% than bare land.The stability of soil moisture content in deep layer was enhanced and the physical properties stability of soil was maintained.The soil moisture content in different depth of soil had significant difference and the changes of soil moisture content were obviously different.[Conclusion]The vegetation coverage of slope could change the soil hydrology obviously and keep soil moisture at the higher level,especially at soil layer below 20 cm.

Spatial and temporal variations of soil moisture in three types of agroforestry boundaries in the Loess Plateau, China

Agroforestry is a ubiquitous landscape on the slopes in Loess Plateau, where soil moisture is a limiting factor for plant growth and development. The spatial and temporal characteristics of soil moisture were studied in three types of agroforestry boundaries： forest-grassland, forest-cropland and shelterbelt-cropland. The result shows that soil moisture content decreased with soil depth increasing from the surface to 110 cm. Soil moisture content differed significantly among the three boundaries all in the rainy season （July-September）, dry season （May-June） and spring （March-April）. The horizontal distribution of soil moisture in different soil layers in the three types of boundaries showed different patterns with line form, wave form, scoop form or ＂W＂ form. The distance of edge influence （DEI） of soil moisture in different types of landscape boundaries was estimated by variance analysis and multiple comparisons. In dry season the DEI in 0-10 cm soil layer was 0.4 H （H, average height of trees）, which ranged from 0.2 H in grassland or in cropland to 0.2 H in forest field for both forest-grassland and forest-cropland boundaries, and 0.7 H （ranged from 0.2 H in cropland to 0.5 H in forest field） for shelterbelt-cropland boundary. In rainy season the DEI at soil depth of 0-110 cm was 0.7 H for the three boundaries. The results indicated that agroforestry type should be carefully selected to maintain soil moisture in land management, especially in restoring degraded land in Loess Plateau.

An overview of soil moisture drought research in China:Progress and perspective

论文回顾了中国土壤湿度干旱(SMD)历史重建和季节预测研究进展,并对未来研究进行了展望,自1950s年代以来,全国整体干旱频率增加,持续时间延长,且有明显区域特征.SMD预测多是利用土壤湿度与气候变量之间的统计关系,而少量基于动力学方法的干旱预测研究强调了初始条件和大气强迫数据对季节尺度干旱预测的重要性,本论文提出:1)加强多时间尺度,跨区域的SMD研究;2)联合气候预测系统,陆面模式和多源土壤湿度数据研制SMD预测系统。

Architectural Plasticity in Response to Soil Moisture in the Stoloniferous Herb, Duchesnea indica

Architectural plasticity of clonal plants may enhance exploitation of soil moisture heterogeneity by the plants. The plasticity of clonal architecture in response to soil moisture in the stoloniferous herb, Duchesnea indica Focke, was investigated in an experiment with different soil moisture contents as treatments, i.e. 40%, 60%, 80%, 100% of the maximum moisture content of soil (MMCS). As soil moisture content increased, the spacer length, ramet density, branching intensity and branching angle of D. indica plants changed by quadratic curve. And the optimum habitat for the plants was at 80% of the MMCS. This architectural plasticity in D. indica was simulated through the Dynamic Logistic Model. The imitative effect was statistically satisfactory. Its architectural plasticity observed here may allow the species to show foraging behavior in its habitat where soil moisture is patchily distributed.

Effect of soil moisture gradient on structure of broad-leaved/Korean pine forest in Changbai Mountain

A 112 m×8 m sample pot which includes 14 sub-plots was set up along the slope in Hongshi Forestry Farm of Baihe Forestry Bureau (127°55′E, 42°30′ N), Jilin Province in August 2002. Community structure, soil moisture contents at 0–10 cm and 10–20 cm in depth, water content of litter as well as the contents of C, N and P of litter, living leaves and branches in the broad-leaved/Korean pine (Pinus korraiensis) forest were measured in each sub-plot on different slope positions. The analytical results showed that there existed an obvious soil moisture gradient along the slope: upper slope <middle slope< lower slope. The difference in soil moisture contents on different positions of slope led to a change of the stand structure of the braod-leaved/Korean pine forest. The proportion ofQuercus mongolica gradually increased with the decrease of soil moisture content and that of other major tree species in the broad-leaved/Korean pine forest gradually decreased or disappeared. The dynamic of soil moisture contents in the litter layer was as same as that in mineral soils. The decomposition rates of the litter on different slope positions were different and the dry weights of existent litter varied significantly. The soil nutrients in the litter on the lower slope was richer than that on the upper slope due to the different stand structure on the different slope positions. The moisture content and nutrient contents of soil had effects on the composition, decomposition, and the nutrient release of litter, thus affecting stands growth and stand structure and finally leading to the change of ecosystem. Key words Soil moisture gradient - nutrient - Stand structure - Broad-leaved/Korean pine forest CLC number S718.5 Document code A Foundation item: This study was supported by the NKBRSF (G1999043407-1), Tackle Key Problem of Science and technology of China (2001BA510B-07), Knowledge Innovation Program of the Chinese Academy of Sciences (KZCX2-406, SCXZD0101), NKTRDP (2001BA510B-07. 2002BA516A20).Biography: WANG Yan (1970-), female, Ph. D, associate professorResponsible editor: Song Funan

Variability of Soil Moisture and Its Relationship with Surface Albedo and Soil Thermal Parameters over the Loess Plateau

Data from July 2006 to June 2008 observed at SACOL （Semi-Arid Climate and Environment Observatory of Lanzhou University, 35.946°N, 104.137°E, elev. 1961 m）, a semi-arid site in Northwest China, are used to study seasonal variability of soil moisture, along with surface albedo and other soil thermal parameters, such as heat capacity, thermal conductivity and thermal diffusivity, and their relationships to soil moisture content. The results indicate that surface albedo decreases with increases in soil moisture content, showing a typical exponential relation between the surface albedo and the soil moisture. The heat capacity, the soil thermal diffusivity, and soil thermal conductivity show large variations between Julian day 90-212 and 450-578. The soil thermal conductivity is found to increase as a power function of soil moisture. Soil heat capacity and soil thermal diffusivity increase with increases in soil moisture. The SACOL observed soil moisture are also used to validate the AMSR-E/AQUA retrieved soil moisture and there is good agreement between them. The analysis of the relationship between satellite retrieved soil moisture and precipitation suggests that the variability of soil moisture depends on the variation of precipitation over the Loess Plateau.

Soil Moisture Variation under Different Irrigation Quota

[Objective] The aim was to study variation of soil moisture under different irrigation quota.[Method] By using Trime-TDR apparatus,soil moisture with different irrigation quota infiltration was measured;combining the characteristics of soil texture,curve characteristics of soil moisture variation with soil depth under different irrigation quota were analyzed.[Result] Different irrigation quota has resulted in variation of soil moisture in different layer depth.Soil moisture is 9.88%,17%,25% and 24.45% in so...

Diagnostic Analysis of Soil Moisture in Northeast China

The horizontal distribution and vertical distribution characteristics of monthly average soil moisture(10-100 cm)of northeastern region of China in 22 years(1981-2002)were analyzed.The spatial and temporal variations also were analyzed.The results showed that from 1981 to 2002,the northeast region surface(10-60 cm)of soil moisture show a downward trend,while the deep(60 cm)following an upward trend.The vertical structure of soil moisture of different years was different.Rainfall was one possible reason for ...