Geoscientific Model Development最新文献

{"title":"Rapid Adaptive Optimization Model for Atmospheric Chemistry (ROMAC) v1.0","authors":"Jiangyong Li, Chunlin Zhang, Wenlong Zhao, Shijie Han, Yu Wang, Hao Wang, Boguang Wang","doi":"10.5194/gmd-16-6049-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-6049-2023","url":null,"abstract":"Abstract. The Rapid Adaptive Optimization Model for Atmospheric Chemistry (ROMAC) is a flexible and computationally efficient photochemical box model. Its unique adaptive dynamic optimization module allows for the dynamic and rapid estimation of the impact of chemical and physical processes on pollutant concentration. ROMAC outperforms traditional box models in evaluating the influence of physical processes on pollutant concentrations. Its ability to quantify the effects of chemical and physical processes on pollutant concentrations has been confirmed through chamber and field observation cases. Since the development of a variable-step and variable-order numerical solver that eliminates the need for Jacobian matrix processing, the computational efficiency of ROMAC has seen a marked improvement with only a marginal increase in error. Specifically, the computational efficiency has improved by 96 % when compared to several established box models, such as F0AM and AtChem. Moreover, the solver maintains a discrepancy of less than 0.1 % when its results are compared with those obtained from a high-precision solver in AtChem.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"80 ","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-30","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136102329","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A standardized methodology for the validation of air quality forecast applications (F-MQO): lessons learnt from its application across Europe","authors":"Lina Vitali, Kees Cuvelier, Antonio Piersanti, Alexandra Monteiro, Mario Adani, Roberta Amorati, Agnieszka Bartocha, Alessandro D'Ausilio, Paweł Durka, Carla Gama, Giulia Giovannini, Stijn Janssen, Tomasz Przybyła, Michele Stortini, Stijn Vranckx, Philippe Thunis","doi":"10.5194/gmd-16-6029-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-6029-2023","url":null,"abstract":"Abstract. A standardized methodology for the validation of short-term air quality forecast applications was developed in the framework of the Forum for Air quality Modeling (FAIRMODE) activities. The proposed approach, focusing on specific features to be checked when evaluating a forecasting application, investigates the model's capability to detect sudden changes in pollutant concentration levels, predict threshold exceedances and reproduce air quality indices. The proposed formulation relies on the definition of specific forecast modelling quality objectives and performance criteria, defining the minimum level of quality to be achieved by a forecasting application when it is used for policy purposes. The persistence model, which uses the most recent observed value as the predicted value, is used as a benchmark for the forecast evaluation. The validation protocol has been applied to several forecasting applications across Europe, using different modelling paradigms and covering a range of geographical contexts and spatial scales. The method is successful, with room for improvement, in highlighting shortcomings and strengths of forecasting applications. This provides a useful basis for using short-term air quality forecasts as a supporting tool for providing correct information to citizens and regulators.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"269 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-27","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136262286","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Application of the Multi-Scale Infrastructure for Chemistry and Aerosols version 0 (MUSICAv0) for air quality research in Africa","authors":"Wenfu Tang, Louisa K. Emmons, Helen M. Worden, Rajesh Kumar, Cenlin He, Benjamin Gaubert, Zhonghua Zheng, Simone Tilmes, Rebecca R. Buchholz, Sara-Eva Martinez-Alonso, Claire Granier, Antonin Soulie, Kathryn McKain, Bruce C. Daube, Jeff Peischl, Chelsea Thompson, Pieternel Levelt","doi":"10.5194/gmd-16-6001-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-6001-2023","url":null,"abstract":"Abstract. The Multi-Scale Infrastructure for Chemistry and Aerosols Version 0 (MUSICAv0) is a new community modeling infrastructure that enables the study of atmospheric composition and chemistry across all relevant scales. We develop a MUSICAv0 grid with Africa refinement (∼ 28 km × 28 km over Africa). We evaluate the MUSICAv0 simulation for 2017 with in situ observations and compare the model results to satellite products over Africa. A simulation from the Weather Research and Forecasting model coupled with Chemistry (WRF-Chem), a regional model that is widely used in Africa studies, is also included in the analyses as a reference. Overall, the performance of MUSICAv0 is comparable to WRF-Chem. Both models underestimate carbon monoxide (CO) compared to in situ observations and satellite CO column retrievals from the Measurements of Pollution in the Troposphere (MOPITT) satellite instrument. MUSICAv0 tends to overestimate ozone (O3), likely due to overestimated stratosphere-to-troposphere flux of ozone. Both models significantly underestimate fine particulate matter (PM2.5) at two surface sites in East Africa. The MUSICAv0 simulation agrees better with aerosol optical depth (AOD) retrievals from the Moderate Resolution Imaging Spectroradiometer (MODIS) and tropospheric nitrogen dioxide (NO2) column retrievals from the Ozone Monitoring Instrument (OMI) than WRF-Chem. MUSICAv0 has a consistently lower tropospheric formaldehyde (HCHO) column than OMI retrievals. Based on model–satellite discrepancies between MUSICAv0 and WRF-Chem and MOPITT CO, MODIS AOD, and OMI tropospheric NO2, we find that future field campaign(s) and more in situ observations in the East African region (5∘ S–5∘ N, 30–45∘ E) could substantially improve the predictive skill of atmospheric chemistry model(s). This suggested focus region exhibits the largest model–in situ observation discrepancies, as well as targets for high population density, land cover variability, and anthropogenic pollution sources.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"47 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134908205","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Overcoming barriers to enable convergence research by integrating ecological and climate sciences: the NCAR–NEON system Version 1","authors":"Danica L. Lombardozzi, William R. Wieder, Negin Sobhani, Gordon B. Bonan, David Durden, Dawn Lenz, Michael SanClements, Samantha Weintraub-Leff, Edward Ayres, Christopher R. Florian, Kyla Dahlin, Sanjiv Kumar, Abigail L. S. Swann, Claire M. Zarakas, Charles Vardeman, Valerio Pascucci","doi":"10.5194/gmd-16-5979-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5979-2023","url":null,"abstract":"Abstract. Global change research demands a convergence among academic disciplines to understand complex changes in Earth system function. Limitations related to data usability and computing infrastructure, however, present barriers to effective use of the research tools needed for this cross-disciplinary collaboration. To address these barriers, we created a computational platform that pairs meteorological data and site-level ecosystem characterizations from the National Ecological Observatory Network (NEON) with the Community Terrestrial System Model (CTSM) that is developed with university partners at the National Center for Atmospheric Research (NCAR). This NCAR–NEON system features a simplified user interface that facilitates access to and use of NEON observations and NCAR models. We present preliminary results that compare observed NEON fluxes with CTSM simulations and describe how the collaboration between NCAR and NEON that can be used by the global change research community improves both the data and model. Beyond datasets and computing, the NCAR–NEON system includes tutorials and visualization tools that facilitate interaction with observational and model datasets and further enable opportunities for teaching and research. By expanding access to data, models, and computing, cyberinfrastructure tools like the NCAR–NEON system will accelerate integration across ecology and climate science disciplines to advance understanding in Earth system science and global change.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"168 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-26","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"136381510","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Evaluation of vertically resolved longwave radiation in SPARTACUS-Urban 0.7.3 and the sensitivity to urban surface temperatures","authors":"Megan A. Stretton, William Morrison, Robin J. Hogan, Sue Grimmond","doi":"10.5194/gmd-16-5931-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5931-2023","url":null,"abstract":"Abstract. Cities' materials and urban form impact radiative exchanges and surface and air temperatures. Here, the SPARTACUS (Speedy Algorithm for Radiative Transfer through Cloud Sides) multi-layer approach to modelling longwave radiation in urban areas (SPARTACUS-Urban) is evaluated using the explicit DART (Discrete Anisotropic Radiative Transfer) model. SPARTACUS-Urban describes realistic 3D urban geometry statistically rather than assuming an infinite street canyon. Longwave flux profiles are compared across an August day for a 2 km × 2 km domain in central London. Simulations are conducted with multiple temperature configurations, including realistic temperature profiles derived from thermal camera observations. The SPARTACUS-Urban model performs well (cf. DART, 2022) when all facets are prescribed a single temperature, with normalised bias errors (nBEs) <2.5 % for downwelling fluxes, and <0.5 % for top-of-canopy upwelling fluxes. Errors are larger (nBE <8 %) for net longwave fluxes from walls and roofs. Using more realistic surface temperatures, varying depending on surface shading, the nBE in upwelling longwave increases to ∼2 %. Errors in roof and wall net longwave fluxes increase through the day, but nBEs are still 8 %–11 %. This increase in nBE occurs because SPARTACUS-Urban represents vertical but not horizontal surface temperature variation within a domain. Additionally, SPARTACUS-Urban outperforms the Harman single-layer canyon approach, particularly in the longwave interception by roofs. We conclude that SPARTACUS-Urban accurately predicts longwave fluxes, requiring less computational time (cf. DART, 2022) but with larger errors when surface temperatures vary due to shading. SPARTACUS-Urban could enhance multi-layer urban energy balance scheme prediction of within-canopy temperatures and fluxes.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135571061","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A Regional multi-Air Pollutant Assimilation System (RAPAS v1.0) for emission estimates: system development and application","authors":"Shuzhuang Feng, Fei Jiang, Zheng Wu, Hengmao Wang, Wei He, Yang Shen, Lingyu Zhang, Yanhua Zheng, Chenxi Lou, Ziqiang Jiang, Weimin Ju","doi":"10.5194/gmd-16-5949-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5949-2023","url":null,"abstract":"Abstract. Top-down atmospheric inversion infers surface–atmosphere fluxes from spatially distributed observations of atmospheric composition in order to quantify anthropogenic and natural emissions. In this study, we developed a Regional multi-Air Pollutant Assimilation System (RAPAS v1.0) based on the Weather Research and Forecasting–Community Multiscale Air Quality (WRF–CMAQ) modeling system model, the three-dimensional variational (3D-Var) algorithm, and the ensemble square root filter (EnSRF) algorithm. This system can simultaneously assimilate hourly in situ CO, SO2, NO2, PM2.5, and PM10 observations to infer gridded emissions of CO, SO2, NOx, primary PM2.5 (PPM2.5), and coarse PM10 (PMC) on a regional scale. In each data assimilation window, we use a “two-step” scheme, in which the emissions are inferred first and then input into the CMAQ model to simulate initial conditions (ICs) of the next window. The posterior emissions are then transferred to the next window as prior emissions, and the original emission inventory is only used in the first window. Additionally, a “super-observation” approach is implemented to decrease the computational costs, observation error correlations, and influence of representative errors. Using this system, we estimated the emissions of CO, SO2, NOx, PPM2.5, and PMC in December and July 2016 over China using nationwide surface observations. The results show that compared to the prior emissions (2016 Multi-resolution Emission Inventory for China – MEIC 2016)), the posterior emissions of CO, SO2, NOx, PPM2.5, and PMC in December 2016 increased by 129 %, 20 %, 5 %, 95 %, and 1045 %, respectively, and the emission uncertainties decreased by 44 %, 45 %, 34 %, 52 %, and 56 %, respectively. With the inverted emissions, the RMSE of simulated concentrations decreased by 40 %–56 %. Sensitivity tests were conducted with different prior emissions, prior uncertainties, and observation errors. The results showed that the two-step scheme employed in RAPAS is robust in estimating emissions using nationwide surface observations over China. This study offers a useful tool for accurately quantifying multi-species anthropogenic emissions at large scales and in near-real time.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"53 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135617074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"A high-resolution marine mercury model MITgcm-ECCO2-Hg with online biogeochemistry","authors":"Siyu Zhu, Peipei Wu, Siyi Zhang, Oliver Jahn, Shu Li, Yanxu Zhang","doi":"10.5194/gmd-16-5915-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5915-2023","url":null,"abstract":"Abstract. Mercury (Hg) is a global persistent contaminant. Modeling studies are useful means of synthesizing a current understanding of the Hg cycle. Previous studies mainly use coarse-resolution models, which makes it impossible to analyze the role of turbulence in the Hg cycle and inaccurately describes the transport of kinetic energy. Furthermore, all of them are coupled with offline biogeochemistry, and therefore they cannot respond to short-term variability in oceanic Hg concentration. In our approach, we utilize a high-resolution ocean model (MITgcm-ECCO2, referred to as “high-resolution-MITgcm”) coupled with the concurrent simulation of biogeochemistry processes from the Darwin Project (referred to as “online”). This integration enables us to comprehensively simulate the global biogeochemical cycle of Hg with a horizontal resolution of 1/5∘. The finer portrayal of surface Hg concentrations in estuarine and coastal areas, strong western boundary flow and upwelling areas, and concentration diffusion as vortex shapes demonstrate the effects of turbulence that are neglected in previous models. Ecological events such as algal blooms can cause a sudden enhancement of phytoplankton biomass and chlorophyll concentrations, which can also result in a dramatic change in particle-bound Hg (HgaqP) sinking flux simultaneously in our simulation. In the global estuary region, including riverine Hg input in the high-resolution model allows us to reveal the outward spread of Hg in an eddy shape driven by fine-scale ocean currents. With faster current velocities and diffusion rates, our model captures the transport and mixing of Hg from river discharge in a more accurate and detailed way and improves our understanding of Hg cycle in the ocean.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"10 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135618822","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Key factors for quantitative precipitation nowcasting using ground weather radar data based on deep learning","authors":"Daehyeon Han, Jungho Im, Yeji Shin, Juhyun Lee","doi":"10.5194/gmd-16-5895-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5895-2023","url":null,"abstract":"Abstract. Quantitative precipitation nowcasting (QPN) can help to reduce the enormous socioeconomic damage caused by extreme weather. The QPN has been a challenging topic due to rapid atmospheric variability. Recent QPN studies have proposed data-driven models using deep learning (DL) and ground weather radar. Previous studies have primarily focused on developing DL models, but other factors for DL-QPN have not been thoroughly investigated. This study examined four critical factors in DL-QPN, focusing on their impact on forecasting performance. These factors are the deep learning model (U-Net, as well as a convolutional long short-term memory, or ConvLSTM), input past sequence length (1, 2, or 3 h), loss function (mean squared error, MSE, or balanced MSE, BMSE), and ensemble aggregation. A total of 24 schemes were designed to measure the effects of each factor using weather radar data from South Korea with a maximum lead time of 2 h. A long-term evaluation was conducted for the summers of 2020–2022 from an operational perspective, and a heavy rainfall event was analyzed to examine an extreme case. In both evaluations, U-Net outperformed ConvLSTM in overall accuracy metrics. For the critical success index (CSI), MSE loss yielded better results for both models in the weak intensity range (≤ 5 mm h−1), whereas BMSE loss was more effective for heavier precipitation. There was a small trend where a longer input time (3 h) gave better results in terms of MSE and BMSE, but this effect was less significant than other factors. The ensemble by averaging results of using MSE and BMSE losses provided balanced performance across all aspects, suggesting a potential strategy to improve skill scores when implemented with optimal weights for each member. All DL-QPN schemes exhibited problems with underestimation and overestimation when trained by MSE and BMSE losses, respectively. All DL models produced blurry results as the lead time increased, while the non-DL model retained detail in prediction. With a comprehensive comparison of these crucial factors, this study offers a modeling strategy for future DL-QPN work using weather radar data.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135617749","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"SedTrace 1.0: a Julia-based framework for generating and running reactive-transport models of marine sediment diagenesis specializing in trace elements and isotopes","authors":"Jianghui Du","doi":"10.5194/gmd-16-5865-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5865-2023","url":null,"abstract":"Abstract. Trace elements and isotopes (TEIs) are important tools in studying ocean biogeochemistry. Understanding their modern ocean budgets and using their sedimentary records to reconstruct paleoceanographic conditions require a mechanistic understanding of the diagenesis of TEIs, yet the lack of appropriate modeling tools has limited our ability to perform such research. Here I introduce SedTrace, a modeling framework that can be used to generate reactive-transport code for modeling marine sediment diagenesis and assist model simulation using advanced numerical tools in Julia. SedTrace enables mechanistic TEI modeling by providing flexible tools for pH and speciation modeling, which are essential in studying TEI diagenesis. SedTrace is designed to solve one particular challenge facing users of diagenetic models: existing models are usually case-specific and not easily adaptable for new problems such that the user has to choose between modifying published code and writing their own code, both of which demand strong coding skills. To lower this barrier, SedTrace can generate diagenetic models only requiring the user to supply Excel spreadsheets containing necessary model information. The resulting code is clearly structured and readable, and it is integrated with Julia's differential equation solving ecosystems, utilizing tools such as automatic differentiation, sparse numerical methods, Newton–Krylov solvers and preconditioners. This allows efficient solution of large systems of stiff diagenetic equations. I demonstrate the capacity of SedTrace using case studies of modeling the diagenesis of pH as well as radiogenic and stable isotopes of TEIs.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"8 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-20","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135617129","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Ensemble of optimised machine learning algorithms for predicting surface soil moisture content at a global scale","authors":"Qianqian Han, Yijian Zeng, Lijie Zhang, Calimanut-Ionut Cira, Egor Prikaziuk, Ting Duan, Chao Wang, Brigitta Szabó, Salvatore Manfreda, Ruodan Zhuang, Bob Su","doi":"10.5194/gmd-16-5825-2023","DOIUrl":"https://doi.org/10.5194/gmd-16-5825-2023","url":null,"abstract":"Abstract. Accurate information on surface soil moisture (SSM) content at a global scale under different climatic conditions is important for hydrological and climatological applications. Machine-learning-based systematic integration of in situ hydrological measurements, complex environmental and climate data, and satellite observation facilitate the generation of reliable data products to monitor and analyse the exchange of water, energy, and carbon in the Earth system at a proper space–time resolution. This study investigates the estimation of daily SSM using 8 optimised machine learning (ML) algorithms and 10 ensemble models (constructed via model bootstrap aggregating techniques and five-fold cross-validation). The algorithmic implementations were trained and tested using International Soil Moisture Network (ISMN) data collected from 1722 stations distributed across the world. The result showed that the K-neighbours Regressor (KNR) had the lowest root-mean-square error (0.0379 cm3 cm−3) on the “test_random” set (for testing the performance of randomly split data during training), the Random Forest Regressor (RFR) had the lowest RMSE (0.0599 cm3 cm−3) on the “test_temporal” set (for testing the performance on the period that was not used in training), and AdaBoost (AB) had the lowest RMSE (0.0786 cm3 cm−3) on the “test_independent-stations” set (for testing the performance on the stations that were not used in training). Independent evaluation on novel stations across different climate zones was conducted. For the optimised ML algorithms, the median RMSE values were below 0.1 cm3 cm−3. GradientBoosting (GB), Multi-layer Perceptron Regressor (MLPR), Stochastic Gradient Descent Regressor (SGDR), and RFR achieved a median r score of 0.6 in 12, 11, 9, and 9 climate zones, respectively, out of 15 climate zones. The performance of ensemble models improved significantly, with the median RMSE value below 0.075 cm3 cm−3 for all climate zones. All voting regressors achieved r scores of above 0.6 in 13 climate zones; BSh (hot semi-arid climate) and BWh (hot desert climate) were the exceptions because of the sparse distribution of training stations. The metric evaluation showed that ensemble models can improve the performance of single ML algorithms and achieve more stable results. Based on the results computed for three different test sets, the ensemble model with KNR, RFR and Extreme Gradient Boosting (XB) performed the best. Overall, our investigation shows that ensemble machine learning algorithms have a greater capability with respect to predicting SSM compared with the optimised or base ML algorithms; this indicates their huge potential applicability in estimating water cycle budgets, managing irrigation, and predicting crop yields.","PeriodicalId":12799,"journal":{"name":"Geoscientific Model Development","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2023-10-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"135779042","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":3,"RegionCategory":"地球科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}