An Intelligent Fine-Tuned Forecasting Technique for Covid-19 Prediction Using Neuralprophet Model

COVID-19,being the virus of fear and anxiety,is one of the most recent and emergent of various respiratory disorders.It is similar to the MERS-COV and SARS-COV,the viruses that affected a large population of different countries in the year 2012 and 2002,respectively.Various standard models have been used for COVID-19 epidemic prediction but they suffered from low accuracy due to lesser data availability and a high level of uncertainty.The proposed approach used a machine learning-based time-series Facebook NeuralProphet model for prediction of the number of death as well as confirmed cases and compared it with Poisson Distribution,and Random Forest Model.The analysis upon dataset has been performed considering the time duration from January 1st 2020 to16th July 2021.The model has been developed to obtain the forecast values till September 2021.This study aimed to determine the pandemic prediction of COVID-19 in the second wave of coronavirus in India using the latest Time-Series model to observe and predict the coronavirus pandemic situation across the country.In India,the cases are rapidly increasing day-by-day since mid of Feb 2021.The prediction of death rate using the proposed model has a good ability to forecast the COVID-19 dataset essentially in the second wave.To empower the prediction for future validation,the proposed model works effectively.

The Altitudinal Belts of Subalpine Virgin Forest on Mt.Gongga Simulated by a Succession Model

How to accurately simulate the distribution of forest species based upon their biological attributes has been a traditional biogeographical issue.Forest gap models are very useful tools for examining the dynamics of forest succession and revealing the species structure of vegetation.In the present study,the GFSM(Gongga Forest Succession Model) was developed and applied to simulate the distribution,composition and succession process of forests in 100 m elevation intervals.The results indicate that the simulated results of the tree species,quantities of the different types of trees,tree age and differences in DBH(diameter at breast height) composition were in line with the actual situation from 1400 to 3700 MASL(meters above sea level) on the eastern slope of Mt.Gongga.Moreover,the dominant species in the simulated results were the same as those in the surveyed database.Thus,the GFSM model can best simulate the features of forest dynamics and structure in the natural conditions of Mt.Gongga.The work provides a new approach to studying the structure and distribution characteristics of mountain ecosystems in varied elevations.Moreover,the results of this study suggest that the biogeochemistry mechanism model should be combined with the forestsuccession model to facilitate the ecological model in simulating the physical and chemical processes involved.

Establishment of models to predict factors influencing periodontitis in patients with type 2 diabetes mellitus

BACKGROUND Type 2 diabetes mellitus(T2DM)is associated with periodontitis.Currently,there are few studies proposing predictive models for periodontitis in patients with T2DM.AIM To determine the factors influencing periodontitis in patients with T2DM by constructing logistic regression and random forest models.METHODS In this a retrospective study,300 patients with T2DM who were hospitalized at the First People’s Hospital of Wenling from January 2022 to June 2022 were selected for inclusion,and their data were collected from hospital records.We used logistic regression to analyze factors associated with periodontitis in patients with T2DM,and random forest and logistic regression prediction models were established.The prediction efficiency of the models was compared using the area under the receiver operating characteristic curve(AUC).RESULTS Of 300 patients with T2DM,224 had periodontitis,with an incidence of 74.67%.Logistic regression analysis showed that age[odds ratio(OR)=1.047,95%confidence interval(CI):1.017-1.078],teeth brushing frequency(OR=4.303,95%CI:2.154-8.599),education level(OR=0.528,95%CI:0.348-0.800),glycosylated hemoglobin(HbA1c)(OR=2.545,95%CI:1.770-3.661),total cholesterol(TC)(OR=2.872,95%CI:1.725-4.781),and triglyceride(TG)(OR=3.306,95%CI:1.019-10.723)influenced the occurrence of periodontitis(P<0.05).The random forest model showed that the most influential variable was HbA1c followed by age,TC,TG, education level, brushing frequency, and sex. Comparison of the prediction effects of the two models showedthat in the training dataset, the AUC of the random forest model was higher than that of the logistic regressionmodel (AUC = 1.000 vs AUC = 0.851;P < 0.05). In the validation dataset, there was no significant difference in AUCbetween the random forest and logistic regression models (AUC = 0.946 vs AUC = 0.915;P > 0.05).CONCLUSION Both random forest and logistic regression models have good predictive value and can accurately predict the riskof periodontitis in patients with T2DM.

Modelling potential distribution of a pine bark beetle in Mexican temperate forests using forecast data and spatial analysis tools

Accurate and reliable predictions of pest species distributions in forest ecosystems are urgently needed by forest managers to develop management plans and monitor new areas of potential establishment.Presence-only species distribution models are commonly used in these evaluations.The maximum entropy algorithm(MaxEnt)has gained popularity for modelling species distribution.Here,MaxEnt was used to model the spatial distribution of the Mexican pine bark beetle(Dendroctonus mexicanus)in a daily fashion by using forecast data from the Weather Research and Forecasting model.This study aimed to exploit freely available geographic and environmental data and software and thus provide a pathway to overcome the lack of costly data and technical guidance that are a challenge to implementing national monitoring and management strategies in developing countries.Our results showed overall agreement values between 60 and 87%.The results of this research can be used for D.mexicanus monitoring and management and may aid as a model to monitor similar species.

Unsuspected implications arising from assumptions in simulations: insights from recasting a forest growth model in system dynamics

Background:Familiarity with a simulation platform can seduce modellers into accepting untested assumptions for convenience of implementation.These assumptions may have consequences greater than commonly suspected,and it is important that modellers remain mindful of assumptions and remain diligent with sensitivity testing.Methods:Familiarity with a technique can lead to complacency,and alternative approaches and software can reveal untested assumptions.Visual modelling environments based on system dynamics may help to make critical assumptions more evident by offering an accessible visual overview and empowering a focus on representational rather than computational efficiency.This capacity is illustrated using a cohort-based forest growth model developed for mixed species forest.Results:The alternative model implementation revealed that untested assumptions in the original model could have substantial influence on simulated outcomes.Conclusions:An important implication is that modellers should remain conscious of all assumptions,consider alternative implementations that reveal assumptions more clearly,and conduct sensitivity tests to inform decisions.

Desertification status mapping in MuttumaWatershed by using Random Forest Model

Potential of the Random Forest Model on mapping of different desertification processes was studied in Muttuma watershed of mid-Murrumbidgee river region of New South Wales,Australia.Desertification vulnerability index was developed using climate,terrain,vegetation,soil and land quality indices to identify environmentally sensitive areas for desertification.Random Forest Model(RFM)was used to predict the different desertification processes such as soil erosion,salinization and waterlogging in the watershed and the information needed to train classification algorithms was obtained from satellite imagery interpretation and ground truth data.Climatic factors(evaporation,rainfall,temperature),terrain factors(aspect,slope,slope length,steepness,and wetness index),soil properties(pH,organic carbon,clay and sand content)and vulnerability indices were used as an explanatory variable.Classification accuracy and kappa index were calculated for training and testing datasets.We recorded an overall accuracy rate of 87.7%and 72.1%for training and testing sites,respectively.We found larger discrepancies between overall accuracy rate and kappa index for testing datasets(72.2%and 27.5%,respectively)suggesting that all the classes are not predicted well.The prediction of soil erosion and no desertification process was good and poor for salinization and water-logging process.Overall,the results observed give a new idea of using the knowledge of desertification process in training areas that can be used to predict the desertification processes at unvisited areas.

Establishment and evaluation of a risk prediction model for gestational diabetes mellitus

BACKGROUND Gestational diabetes mellitus(GDM)is a condition characterized by high blood sugar levels during pregnancy.The prevalence of GDM is on the rise globally,and this trend is particularly evident in China,which has emerged as a significant issue impacting the well-being of expectant mothers and their fetuses.Identifying and addressing GDM in a timely manner is crucial for maintaining the health of both expectant mothers and their developing fetuses.Therefore,this study aims to establish a risk prediction model for GDM and explore the effects of serum ferritin,blood glucose,and body mass index(BMI)on the occurrence of GDM.AIM To develop a risk prediction model to analyze factors leading to GDM,and evaluate its efficiency for early prevention.METHODS The clinical data of 406 pregnant women who underwent routine prenatal examination in Fujian Maternity and Child Health Hospital from April 2020 to December 2022 were retrospectively analyzed.According to whether GDM occurred,they were divided into two groups to analyze the related factors affecting GDM.Then,according to the weight of the relevant risk factors,the training set and the verification set were divided at a ratio of 7:3.Subsequently,a risk prediction model was established using logistic regression and random forest models,and the model was evaluated and verified.RESULTS Pre-pregnancy BMI,previous history of GDM or macrosomia,hypertension,hemoglobin(Hb)level,triglyceride level,family history of diabetes,serum ferritin,and fasting blood glucose levels during early pregnancy were determined.These factors were found to have a significant impact on the development of GDM(P<0.05).According to the nomogram model’s prediction of GDM in pregnancy,the area under the curve(AUC)was determined to be 0.883[95%confidence interval(CI):0.846-0.921],and the sensitivity and specificity were 74.1%and 87.6%,respectively.The top five variables in the random forest model for predicting the occurrence of GDM were serum ferritin,fasting blood glucose in early pregnancy,pre-pregnancy BMI,Hb level and triglyceride level.The random forest model achieved an AUC of 0.950(95%CI:0.927-0.973),the sensitivity was 84.8%,and the specificity was 91.4%.The Delong test showed that the AUC value of the random forest model was higher than that of the decision tree model(P<0.05).CONCLUSION The random forest model is superior to the nomogram model in predicting the risk of GDM.This method is helpful for early diagnosis and appropriate intervention of GDM.

Estimating the carbon fluxes of forests with an individual-based forest model

Background： Capturing the response of forest ecosystems to inter-annual climate variability is a great challenge.In this study, we tested the capability of an individual-based forest gap model to display carbon fluxes at yearly and daily time scales.The forest model was applied to a spruce forest to simulate the gross primary production（GPP）, respiration and net ecosystem exchange（NEE）.We analyzed how the variability in climate affected simulated carbon fluxes at the scale of the forest model.Results： Six years were simulated at a daily time scale and compared to the observed eddy covariance（EC） data.In general, the seasonal cycle of the individual carbon fluxes was correctly described by the forest model.However, the estimated GPP differed from the observed data on the days of extreme climatic conditions.Two new parameterizations were developed： one resulting from a numerical calibration, and the other resulting from a filtering method.We suggest new parameter values and even a new function for the temperature limitation of photosynthesis.Conclusions： The forest model reproduced the observed carbon fluxes of a forest ecosystem quite wel.Of the three parameterizations, the calibrated model version performed best.However, the filtering approach showed that calibrated parameter values do not necessarily correctly display the individual functional relations.The concept of simulating forest dynamics at the individual base is a valuable tool for simulating the NEE, GPP and respiration of forest ecosystems.

The carbon fluxes in different successional stages:modelling the dynamics of tropical montane forests in South Ecuador

Background： Tropical forests play an important role in the global carbon（C） cycle.However, tropical montane forests have been studied less than tropical lowland forests, and their role in carbon storage is not well understood.Montane forests are highly endangered due to logging, land-use and climate change.Our objective was to analyse how the carbon balance changes during forest succession.Methods： In this study, we used a method to estimate local carbon balances that combined forest inventory data with process-based forest models.We utilised such a forest model to study the carbon balance of a tropical montane forest in South Ecuador, comparing two topographical slope positions（ravines and lower slopes vs upper slopes and ridges）.Results： The simulation results showed that the forest acts as a carbon sink with a maximum net ecosystem exchange（NEE） of 9.3 Mg C？（ha？yr）-1during its early successional stage（0–100 years）.In the late successional stage, the simulated NEE fluctuated around zero and had a variation of 0.77 Mg C？（ha？yr）–1.The simulated variability of the NEE was within the range of the field data.We discovered several forest attributes（e.g., basal area or the relative amount of pioneer trees） that can serve as predictors for NEE for young forest stands（0–100 years） but not for those in the late successional stage（500–1,000 years）.In case of young forest stands these correlations are high, especially between stand basal area and NEE.Conclusion： In this study, we used an Ecuadorian study site as an example of how to successfully link a forest model with forest inventory data, for estimating stem-diameter distributions, biomass and aboveground net primary productivity.To conclude, this study shows that process-based forest models can be used to investigate the carbon balance of tropical montane forests.With this model it is possible to find hidden relationships between forest attributes and forest carbon fluxes.These relationships promote a better understanding of the role of tropical montane forests in the context of global carbon cycle, which in future will become more relevant to a society under global change.

Observations and Modeling of Incoming Longwave Radiation to Snow Beneath Forest Canopies in the West Tianshan Mountains, China

Forest canopy reduces shortwave radiation and increases the incoming longwave radiation to snowpacks beneath forest canopies. Furthermore, the effect of forest canopy may be changed by complex topography. In this paper, we measured and simulated the incoming longwave radiation to snow beneath forest at different canopy openness in the west Tianshan Mountains, China(43°16'N, 84°24'E) during spring 2013. A sensitivity study was conducted to explore the way that terrain influenced the incoming longwave radiation to snow beneath forest canopies. In the simulation model, measurement datasets, including air temperature, incoming shortwave radiation above canopy, and longwave radiation enhanced by adjacent terrain, were applied to calculate the incoming longwave radiation to snow beneath forest canopy. The simulation results were consistent with the measurements on hourly scale and daily scale. The effect of longwave radiation enhanced by terrain was important than that of shortwave radiation above forest canopy with different openness except the 20% canopy openness. The longwave radiation enhanced due to adjacent terrain increases with the slope increase and temperature rise. When air temperature(or slope) is relatively low, thelongwave radiation enhanced by adjacent terrain is not sensitive to slope(or air temperature), but the sensitivity increases with the decrease of snow cover area on sunny slope. The effect of longwave radiation is especially sensitive when the snow cover on sunny slope melts completely. The effect of incoming shortwave radiation reflected by adjacent terrain on incoming longwave radiation to snow beneath forest canopies is more slight than that of the enhanced longwave radiation.

Traffic flow prediction of urban road network based on LSTM-RF model

Traffic flow prediction,as the basis of signal coordination and travel time prediction,has become a research point in the field of transportation.For traffic flow prediction,researchers have proposed a variety of methods,but most of these methods only use the time domain information of traffic flow data to predict the traffic flow,ignoring the impact of spatial correlation on the prediction of target road segment flow,which leads to poor prediction accuracy.In this paper,a traffic flow prediction model called as long short time memory and random forest(LSTM-RF)was proposed based on the combination model.In the process of traffic flow prediction,the long short time memory(LSTM)model was used to extract the time sequence features of the predicted target road segment.Then,the predicted value of LSTM and the collected information of adjacent upstream and downstream sections were simultaneously used as the input features of the random forest model to analyze the spatial-temporal correlation of traffic flow,so as to obtain the final prediction results.The traffic flow data of 132 urban road sections collected by the license plate recognition system in Guiyang City were tested and verified.The results show that the method is better than the single model in prediction accuracy,and the prediction error is obviously reduced compared with the single model.

Comparison of stem taper models for the four tropical tree species in Mount Makiling, Philippines

This study was conducted to evaluate the performance of six stem taper models on four tropical tree species, namely Celtis luzonica(Magabuyo),Diplodiscus paniculatus(Balobo), Parashorea malaanonan(Bagtikan), and Swietenia macrophylla(Mahogany) in Mount Makiling Forest Reserve(MMFR), Philippines using fit statistics and lack-of-fit statistics. Four statistical criteria were used in this study, including the standard error of estimate(SEE),coefficient of determination(R^2), mean bias( E),and absolute mean difference(AMD). For the lack-offit statistics, SEE, E and AMD were determined in different relative height classes. The results indicated that the Kozak02 stem taper model offered the best fit for the four tropical species in most statistics. The Kozak02 model also consistently provided the best performance in the lack-of-fit statistics with the best SEE, E and AMD in most of the relative height classes. These stem taper equations could help forest managers and researchers better estimate the diameter of the outside bark with any given height,merchantable stem volumes and total stem volumes of standing trees belonging to the four species of thetropical forest in MMFR.

Modeling of forest soil and litter health using disturbance and landscape heterogeneity indicators in northern Iran

This paper focuses on the indicators of soil and litter health, disturbance, and landscape heterogeneity as a tool for prediction of ecosystem sustainability in the northern forests of Iran. The study area was divided into spatial homogenous sites using slope, aspect, and soil humidity classes. Then a range of sites along the disturbance gradient was selected for sampling. Chemical and physical indicators of soil and litter health were measured at random points within these sites. Structural equation modeling(SEM) was applied to link six constructs of landscape heterogeneity, three constructs of disturbance(harvest, livestock, and human accessibility), and soil and litter health. The results showed that with decreasing accessibility, the total N and organic matter content of soil increased and effective bulk density decreased. Harvesting activities increased soil organic matter. Therefore, it is concluded that disturbances through harvesting and accessibility inversely affect the soil health. Unexpectedly, it was found that the litter total C and C:N ratio improved with an increase in the harvest and accessibility disturbances, whereas litter bulk density decreased. Investigation of tree composition revealed that in the climax communities, which are normally affected more by harvesting activities, some species like Fagus orientalis Lipsky with low decomposition rate are dominant. The research results showed that changes in disturbance intensity are reflected in litter and soil indicators, whereas the SEM indicated that landscape heterogeneity has a moderator effect on the disturbance to both litter and soil paths.

Identification of Mixtures of Two Types of Body Fluids Using the Multiplex Methylation System and Random Forest Models

Objective Body fluid mixtures are complex biological samples that frequently occur in crime scenes,and can provide important clues for criminal case analysis.DNA methylation assay has been applied in the identification of human body fluids,and has exhibited excellent performance in predicting single-source body fluids.The present study aims to develop a methylation SNaPshot multiplex system for body fluid identification,and accurately predict the mixture samples.In addition,the value of DNA methylation in the prediction of body fluid mixtures was further explored.Methods In the present study,420 samples of body fluid mixtures and 250 samples of single body fluids were tested using an optimized multiplex methylation system.Each kind of body fluid sample presented the specific methylation profiles of the 10 markers.Results Significant differences in methylation levels were observed between the mixtures and single body fluids.For all kinds of mixtures,the Spearman’s correlation analysis revealed a significantly strong correlation between the methylation levels and component proportions(1:20,1:10,1:5,1:1,5:1,10:1 and 20:1).Two random forest classification models were trained for the prediction of mixture types and the prediction of the mixture proportion of 2 components,based on the methylation levels of 10 markers.For the mixture prediction,Model-1 presented outstanding prediction accuracy,which reached up to 99.3%in 427 training samples,and had a remarkable accuracy of 100%in 243 independent test samples.For the mixture proportion prediction,Model-2 demonstrated an excellent accuracy of 98.8%in 252 training samples,and 98.2%in 168 independent test samples.The total prediction accuracy reached 99.3%for body fluid mixtures and 98.6%for the mixture proportions.Conclusion These results indicate the excellent capability and powerful value of the multiplex methylation system in the identification of forensic body fluid mixtures.

Impact of regional human capital and urban sprawl on forest——based on a forest conservation model

Metropolitan cities in China have become a major economic hubs with an unprecedented increase of land use and decline of environmental resources. Based on a simple and abstract forest conservation model, this paper attempts to explain changes of forest resources caused by urban sprawl. Through the research, it is found that high level of regional human capital is beneficial to curb urban sprawl. In this vein the model presents the urban forest conservation cost strategy at the Nash equilibrium of varied discount factor and parameter control.

Modelling the dead fuel moisture content in a grassland of Ergun City,China

The dead fuel moisture content(DFMC)is the key driver leading to fire occurrence.Accurately estimating the DFMC could help identify locations facing fire risks,prioritise areas for fire monitoring,and facilitate timely deployment of fire-suppression resources.In this study,the DFMC and environmental variables,including air temperature,relative humidity,wind speed,solar radiation,rainfall,atmospheric pressure,soil temperature,and soil humidity,were simultaneously measured in a grassland of Ergun City,Inner Mongolia Autonomous Region of China in 2021.We chose three regression models,i.e.,random forest(RF)model,extreme gradient boosting(XGB)model,and boosted regression tree(BRT)model,to model the seasonal DFMC according to the data collected.To ensure accuracy,we added time-lag variables of 3 d to the models.The results showed that the RF model had the best fitting effect with an R2value of 0.847 and a prediction accuracy with a mean absolute error score of 4.764%among the three models.The accuracies of the models in spring and autumn were higher than those in the other two seasons.In addition,different seasons had different key influencing factors,and the degree of influence of these factors on the DFMC changed with time lags.Moreover,time-lag variables within 44 h clearly improved the fitting effect and prediction accuracy,indicating that environmental conditions within approximately 48 h greatly influence the DFMC.This study highlights the importance of considering 48 h time-lagged variables when predicting the DFMC of grassland fuels and mapping grassland fire risks based on the DFMC to help locate high-priority areas for grassland fire monitoring and prevention.

The Mathematical Model of Short-Term Forest Fire Spread

In this paper, we establish a mathematical model of the forest fire spread process based on a partial differential equation. We describe the distribution of time field and velocity field in the whole two-dimensional space by vector field theory. And we obtain a continuous algorithm to predict the dynamic behavior of forest fire spread in a short time. We use the algorithm to interpolate the fire boundary by cubic non-uniform rational B-spline closed curve. The fire boundary curve at any time can be simulated by solving the Eikonal equation. The model is tested in theory and in practice. The results show that the model has good accuracy and stability, and it’s compatible with most of the existing models, such as the elliptic model and the cellular automata model.

Model-based estimation of above-ground biomass in the miombo ecoregion of Zambia

Background：Information on above-ground biomass（AGB） is important for managing forest resource use at local levels,land management planning at regional levels,and carbon emissions reporting at national and international levels.In many tropical developing countries,this information may be unreliable or at a scale too coarse for use at local levels.There is a vital need to provide estimates of AGB with quantifiable uncertainty that can facilitate land use management and policy development improvements.Model-based methods provide an efficient framework to estimate AGB.Methods：Using National Forest Inventory（NFI） data for a^1,000,000 ha study area in the miombo ecoregion,Zambia,we estimated AGB using predicted canopy cover,environmental data,disturbance data,and Landsat 8 OLI satellite imagery.We assessed different combinations of these datasets using three models,a semiparametric generalized additive model（GAM） and two nonlinear models（sigmoidal and exponential）,employing a genetic algorithm for variable selection that minimized root mean square prediction error（RMSPE）,calculated through cross-validation.We compared model fit statistics to a null model as a baseline estimation method.Using bootstrap resampling methods,we calculated 95% confidence intervals for each model and compared results to a simple estimate of mean AGB from the NFI ground plot data.Results：Canopy cover,soil moisture,and vegetation indices were consistently selected as predictor variables.The sigmoidal model and the GAM performed similarly;for both models the RMSPE was -36.8 tonnes per hectare（i.e.,57% of the mean）.However,the sigmoidal model was approximately 30% more efficient than the GAM,assessed using bootstrapped variance estimates relative to a null model.After selecting the sigmoidal model,we estimated total AGB for the study area at 64,526,209 tonnes（＋/- 477,730）,with a confidence interval 20 times more precise than a simple designbased estimate.Conclusions：Our findings demonstrate that NFI data may be combined with freely available satellite imagery and soils data to estimate total AGB with quantifiable uncertainty,while also providing spatially explicit AGB maps useful for management,planning,and reporting purposes.

Advances in artificial intelligence for predicting complication risks post-laparoscopic radical gastrectomy for gastric cancer:A significant leap forward

In a recent paper,Hong et al developed an artificial intelligence(AI)-driven predictive scoring system for potential complications following laparoscopic radical gastrectomy for gastric cancer patients.They demonstrated that integrating AI with random forest models significantly improved the preoperative prediction and patient outcome management accuracy.By incorporating data from multiple centers,their model ensures standardization,reliability,and broad applicability,distinguishing it from the prior models.The present study highlights AI's potential in clinical decision support,aiding in the preoperative and postoperative management of gastric cancer patients.Our findings may pave the way for future prospective studies to further enhance AI-supported diagnoses in clinical practice.

Understanding and simulating of three-dimensional subsurface hydrological partitioning in an alpine mountainous area, China

Critical zone(CZ)plays a vital role in sustaining biodiversity and humanity.However,flux quantification within CZ,particularly in terms of subsurface hydrological partitioning,remains a significant challenge.This study focused on quantifying subsurface hydrological partitioning,specifically in an alpine mountainous area,and highlighted the important role of lateral flow during this process.Precipitation was usually classified as two parts into the soil:increased soil water content(SWC)and lateral flow out of the soil pit.It was found that 65%–88%precipitation contributed to lateral flow.The second common partitioning class showed an increase in SWC caused by both precipitation and lateral flow into the soil pit.In this case,lateral flow contributed to the SWC increase ranging from 43%to 74%,which was notably larger than the SWC increase caused by precipitation.On alpine meadows,lateral flow from the soil pit occurred when the shallow soil was wetter than the field capacity.This result highlighted the need for three-dimensional simulation between soil layers in Earth system models(ESMs).During evapotranspiration process,significant differences were observed in the classification of subsurface hydrological partitioning among different vegetation types.Due to tangled and aggregated fine roots in the surface soil on alpine meadows,the majority of subsurface responses involved lateral flow,which provided 98%–100%of evapotranspiration(ET).On grassland,there was a high probability(0.87),which ET was entirely provided by lateral flow.The main reason for underestimating transpiration through soil water dynamics in previous research was the neglect of lateral root water uptake.Furthermore,there was a probability of 0.12,which ET was entirely provided by SWC decrease on grassland.In this case,there was a high probability(0.98)that soil water responses only occurred at layer 2(10–20 cm),because grass roots mainly distributed in this soil layer,and grasses often used their deep roots for water uptake during ET.To improve the estimation of soil water dynamics and ET,we established a random forest(RF)model to simulate lateral flow and then corrected the community land model(CLM).RF model demonstrated good performance and led to significant improvements in CLM simulation.These findings enhance our understanding of subsurface hydrological partitioning and emphasize the importance of considering lateral flow in ESMs and hydrological research.