Soil and Tillage Research最新文献

{"title":"Recognizing and reducing effects of moisture-salt coexistence on soil organic matter spectral prediction：From laboratory to satellite","authors":"Danyang Wang, Yayi Tan, Cheng Li, Jingda Xin, Yunqi Wang, Huagang Hou, Lulu Gao, Changbo Zhong, Jianjun Pan, Zhaofu Li","doi":"10.1016/j.still.2024.106397","DOIUrl":"https://doi.org/10.1016/j.still.2024.106397","url":null,"abstract":"Soil organic matter (SOM) mapping in salinized areas is crucial for scientific guidance on soil salinization. However, accurately mapping SOM is challenging due to the intricate interplay between soil moisture content (SMC) and soil salt content (SSC), which significantly influences soil spectra. Unlike prior research that has separately examined the impacts of moisture or salinity, this study delves into the combined effects of these factors on SOM spectra. The objective of this study is to develop and validate several spectral optimization algorithms at both the laboratory and satellite levels. In October 2020, a study was conducted using 291 ground-truth data to examine the impact of various moisture-salt stages (seven moisture stages, five salt stages) on hyperspectral data. Spectrum mechanism responsive for soil moisture and salinity were analyzed through spectral curves, correlation, and analysis of variance (Anova, AOV), and spectrum mechanism responsive for soil moisture and salinity model were built. Following that, the spectra were optimized using piecewise direct normalization (PDS)-AOV, non-negative matrix factorization (NMF)-AOV, and orthogonal signal correction (OSC)-AOV. The SOM prediction models were then built by integrating these optimized spectra with Stacking ensemble machine learning algorithms (RF, GBM, ANN). Eventually, the lab-optimized spectra were merged with satellite multispectral images to create new image (named REC) for SOM mapping. The results indicated varying impacts of SMC and SSC on spectra, particularly between 1400 nm to 2000 nm, revealing the influence of moisture-salt interaction; the best optimization algorithm (OSC-AOV) with Stacking mitigated the effect of moisture-salt coexistence on spectra (the R<ce:sup loc=\"post\">2</ce:sup> and RPD of the best models elevated by 0.005–0.267, 0.020–0.374 respectively, RMSE reduced by 0.137–1.817 g/kg); implementing this algorithm on REC significantly improved the accuracy of SOM mapping (R<ce:sup loc=\"post\">2</ce:sup> elevated by 0.185–0.259, RMSE reduced by 2.615–3.203 g/kg). This study extensively investigated the effects of moisture and salinity on spectra, spanning from laboratory to satellite, offering a novel approach to understanding and addressing the complexities in SOM mapping in salinized environments.","PeriodicalId":501007,"journal":{"name":"Soil and Tillage Research","volume":"34 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-12-05","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"142788902","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Simulating field soil temperature variations with physics-informed neural networks","authors":"Xiaoting Xie, Hengnian Yan, Yili Lu, Lingzao Zeng","doi":"10.1016/j.still.2024.106236","DOIUrl":"https://doi.org/10.1016/j.still.2024.106236","url":null,"abstract":"Information on soil temperature is crucial for modeling hydrological and climatic processes. Nevertheless, direct measurements of soil temperature are usually rather limited in space, leading to an urgent need for improved spatial resolution. To address this issue, a Physics-Informed Neural Networks (PINN) method for estimating soil temperature () profile variations was proposed in this study. This method combines the advantages of Deep Neural Networks (DNN) in modeling complex non-linear relationships and physical laws for more robust predictions. The performance was evaluated using in-situ annual soil at depths of 5 cm, 10 cm and 20 cm on a maize field in Northeast China. Cross-validation was used, a PINN was used to derive the new data at unobserved depth from observations at the other two depths. The results demonstrated that the performance of the PINN was superior to the commonly used process-based method and a DNN for all situations. Compared to the traditional method, the PINN achieved a 0.69°C and 0.39°C reduction in root-mean-square error (RMSE) for estimates at 10 cm and 20 cm depths, respectively, under plowed tillage condition, while it could also accurately estimate at 5 cm depth with RMSE of 0.56 °C. In addition, the PINN does not require inputs of soil thermal properties e.g., apparent thermal diffusivity (κ), as the space and time-dependent κ values could also be learned during the training process. The results presented here demonstrated that a PINN could successfully utilize limited observation data to estimate unknown soil profiles, and solve some challenging problems beyond the reach of existing methods in simulating soil thermal dynamics.","PeriodicalId":501007,"journal":{"name":"Soil and Tillage Research","volume":"2 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2024-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"141768699","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}