Dataset Information

Hyperspectral band selection and modeling of soil organic matter content in a forest using the Ranger algorithm.

ABSTRACT: Effective soil spectral band selection and modeling methods can improve modeling accuracy. To establish a hyperspectral prediction model of soil organic matter (SOM) content, this study investigated a forested Eucalyptus plantation in Huangmian Forest Farm, Guangxi, China. The Ranger and Lasso algorithms were used to screen spectral bands. Subsequently, models were established using four algorithms: partial least squares regression, random forest (RF), a support vector machine, and an artificial neural network (ANN). The optimal model was then selected. The results showed that the modeling accuracy was higher when band selection was based on the Ranger algorithm than when it was based on the Lasso algorithm. ANN modeling had the best goodness of fit, and the model established by RF had the most stable modeling results. Based on the above results, a new method is proposed in this study for band selection in the early phase of soil hyperspectral modeling. The Ranger algorithm can be applied to screen the spectral bands, and ANN or RF can then be selected to construct the prediction model based on different datasets, which is applicable to establish the prediction model of SOM content in red soil plantations. This study provides a reference for the remote sensing of soil fertility in forests of different soil types and a theoretical basis for developing portable equipment for the hyperspectral measurement of SOM content in forest habitats.

SUBMITTER: Shi Y

PROVIDER: S-EPMC8238212 | biostudies-literature |

REPOSITORIES: biostudies-literature

ACCESS DATA

Similar Datasets

Project description:Soil moisture, soil organic carbon, and nitrogen content prediction are considered significant fields of study as they are directly related to plant health and food production. Direct estimation of these soil properties with traditional methods, for example, the oven-drying technique and chemical analysis, is a time and resource-consuming approach and can predict only smaller areas. With the significant development of remote sensing and hyperspectral (HS) imaging technologies, soil moisture, carbon, and nitrogen can be estimated over vast areas. This paper presents a generalized approach to predicting three different essential soil contents using a comprehensive study of various machine learning (ML) models by considering the dimensional reduction in feature spaces. In this study, we have used three popular benchmark HS datasets captured in Germany and Sweden. The efficacy of different ML algorithms is evaluated to predict soil content, and significant improvement is obtained when a specific range of bands is selected. The performance of ML models is further improved by applying principal component analysis (PCA), a dimensional reduction method that works with an unsupervised learning method. The effect of soil temperature on soil moisture prediction is evaluated in this study, and the results show that when the soil temperature is considered with the HS band, the soil moisture prediction accuracy does not improve. However, the combined effect of band selection and feature transformation using PCA significantly enhances the prediction accuracy for soil moisture, carbon, and nitrogen content. This study represents a comprehensive analysis of a wide range of established ML regression models using data preprocessing, effective band selection, and data dimension reduction and attempt to understand which feature combinations provide the best accuracy. The outcomes of several ML models are verified with validation techniques and the best- and worst-case scenarios in terms of soil content are noted. The proposed approach outperforms existing estimation techniques.

Project description:Mineral-associated organic matter is associated with a suite of soil minerals that can confer stability, resulting in the potential for long-term storage of carbon (C). Not all interactions impart the same level of protection, however; evidence is suggesting that C in certain mineral pools is dynamic and vulnerable to disturbance in the decades following harvesting. The objective of this research was to describe and characterize organic matter-mineral interactions through depth in horizons of soils of contrasting stand age. Sequential selective dissolutions representing increasingly stable mineral-associated organic matter pools from water soluble minerals (deionized water), organo-metal complexes (Na-pyrophosphate), poorly-crystalline minerals (HCl hydroxylamine), and crystalline secondary minerals (Na-dithionite HCl)) were carried out for Ae, Bf and BC horizons sampled from a Young and Mature forest site (35 and 110 years post-harvest) in Mooseland, Nova Scotia, Canada. Sequential selective dissolution extracts were analyzed for C, ?13C, iron (Fe) and aluminum (Al). Organo-metal complexes (OMC) were the largest mineral-associated OM pool in all horizons. This pool dominated the C distribution in B horizons (~60-70% of Bf bulk C), with a minor contribution from poorly-crystalline (PCrys), crystalline (Crys) minerals and water soluble (WS) associations. C in OMC and PCrys pools explained the variation in bulk C in horizons through depth at both sites. Twice as much C in OMC pools was measured at the Mature site compared to the Young site in the Bf horizons, supported by higher C:(Fe+Al) ratios. Isotopic analysis indicated that this extraction procedure isolated distinct mineral-associated OM pools. ?13C signatures of pyrophosphate-extracted OMC pools ranged from -27‰ to -28‰, similar to ?13C of bulk C and to plant-derived humic acids and associated biomass. The water soluble phase (mean ?13C = -29 ‰) was up to 2 ‰ more depleted, whereas the ?13C of Crys pools were more enriched in 13C (-13‰ to -16 ‰) compared to bulk soil. The results from this study suggest that association with minerals does not necessarily confer stability: organo-metal pools dominate in podzol horizons through depth, and contribute most to C storage, but are potentially susceptible to destabilization following the physical changes resulting from forest harvesting disturbance.

Project description:Rising temperatures enhance microbial decomposition of soil organic matter (SOM) and thereby increase the soil CO2 efflux. Elevated decomposition rates might differently affect distinct SOM pools, depending on their stability and accessibility. Soil fractions derived from density fractionation have been suggested to represent SOM pools with different turnover times and stability against microbial decomposition. To investigate the effect of soil warming on functionally different soil organic matter pools, we here investigated the chemical and isotopic composition of bulk soil and three density fractions (free particulate organic matter, fPOM; occluded particulate organic matter, oPOM; and mineral associated organic matter, MaOM) of a C-rich soil from a long-term warming experiment in a spruce forest in the Austrian Alps. At the time of sampling, the soil in this experiment had been warmed during the snow-free period for seven consecutive years. During that time no thermal adaptation of the microbial community could be identified and CO2 release from the soil continued to be elevated by the warming treatment. Our results, which included organic carbon content, total nitrogen content, ?13C, ?14C, ?15N and the chemical composition, identified by pyrolysis-GC/MS, showed no significant differences in bulk soil between warming treatment and control. Surprisingly, the differences in the three density fractions were mostly small and the direction of warming induced change was variable with fraction and soil depth. Warming led to reduced N content in topsoil oPOM and subsoil fPOM and to reduced relative abundance of N-bearing compounds in subsoil MaOM. Further, warming increased the ?13C of MaOM at both sampling depths, reduced the relative abundance of carbohydrates while it increased the relative abundance of lignins in subsoil oPOM. As the size of the functionally different SOM pools did not significantly change, we assume that the few and small modifications in SOM chemistry result from an interplay of enhanced microbial decomposition of SOM and increased root litter input in the warmed plots. Overall, stable functional SOM pool sizes indicate that soil warming had similarly affected easily decomposable and stabilized SOM of this C-rich forest soil.

Project description:Annual estimates of CO2 and dissolved carbon concentrations in the soil profile provide valuable insight into the dynamics of organic matter in soil and the effect of changes to vegetation cover. The aim of this study was to observe the spatial influence of litter decomposition in the first few centimeters of the soil for CO2 fluxes and to describe the processing of soil organic matter throughout the soil profile by comparing three small tropical watersheds. Data were collected biweekly for six months, from December 2015 to May 2016. CO2 was measured using an infrared gas analyzer in fixed chambers and the dissolved carbon of soil solution was analyzed in a TOC analyzer. No differences were found in the total soil CO2 fluxes (control flux treatments) between the three study areas. In both cacao agroforestry systems (managed and unmanaged), total CO2 fluxes were influenced by the decomposition of litter. CO2 emissions in the soil profile of the cacao agroforestry systems were highly variable, compared to the preserved forest, and highly dependent on the soil characteristics attributed to the type of vegetation cover. Although a definite pattern between the temperature and soil moisture was not identified, these parameters showed a strong relationship in controlling the release of CO2 between treatments. The organic and inorganic dissolved carbon patterns in the soil solution of the three areas revealed different responses of soil organic matter processing related to soil characteristics and vegetation. The results confirm the hypothesis that the top of soils (total CO2 fluxes) of both cacao agroforestry systems (managed and unmanaged) emits fluxes of CO2, which do not differ statistically from the preserved forest. However, depending on the soil characteristics, the cacao agroforestry system can result in an accumulation of CO2 and dissolved inorganic carbon in the soil profile that is prone to being transported by hydrological routes to groundwater and stream water.

Dataset Information

Hyperspectral band selection and modeling of soil organic matter content in a forest using the Ranger algorithm.

Similar Datasets

OmicsDI is part of the ELIXIR infrastructure

Tweets