Improving the remote estimation of soil organic carbon in complex ecosystems with Sentinel-2 and GIS using Gaussian processes regression

Izurieta, Johanna Elizabeth Ayala; Santillán, Carlos Arturo Jara; Márquez, Carmen O.; García, Vega; Rivera-Caicedo, Juan Pablo; Wittenberghe, Shari Van; Delegido, Jesús; Verrelst, Jochem

doi:10.1007/s11104-022-05506-1

Cited by 23 publications

(12 citation statements)

References 104 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The results of lithologies of the intervened area and lower than the median of the intervened area are in concordance with other studies related to SOC on paramo region in Colombia where a positive asymmetry of 0.97 and kurtosis of 5.18, also estimated SOC values that oscillate in the range of 22 and 338 Mg/ha and medians of up to 180.52 Mg/ha are indicated [48]. Also, studies in the herbaceus páramo ecosystem of Ecuador showed values of SOC between 67-221 in the 0-30 cm soil profile and values of SOC between 32 and 179 Mg/ha in the 30-60 cm profile [49].…”

Section: Discussionmentioning

confidence: 83%

Evaluation of Soil Organic Carbon Storage of Atillo in the Ecuadorian Andean Wetlands

et al. 2022

Self Cite

View full text Add to dashboard Cite

Identifying the SOC levels and revealing the potential of SOC storage of ecosystems difficult to sample and study are necessary contributions to the understanding of the global reserves of SOC. Wetlands store large amounts of SOC within their soils. They have an important role in water regulation and have great biological and floristic diversity. Therefore, this study aimed to assess the SOC stock in Atillo micro-watershed in the Ecuadorian Andean wetlands at two soil depths (0–30 cm and 30–60 cm below ground) and to assess the importance of the ecosystem and its conservation in favor of reducing emissions due to degradation processes. For that, we sampled the study zone with 101 composite samples of soil to obtain the SOC storage for each sample point in Mg/ha. A SOC estimation to evaluate its spatial distribution was performed using the geostatistical method Kriging. The results show a high storage capacity of the study zone with SOC values of 126 to 454 Mg/ha in the 0–30 cm soil profile and 148 to 350 Mg/ha in the 30–60 cm soil profile. The preservation and protection mechanisms of high SOC reserves should be taken into account to prevent the emission of CO2.

show abstract

Section: Discussionmentioning

confidence: 83%

Evaluation of Soil Organic Carbon Storage of Atillo in the Ecuadorian Andean Wetlands

et al. 2022

Self Cite

View full text Add to dashboard Cite

show abstract

“…An interesting property of GPC is the ability to provide band relevance, obtained through automatic relevance determination (ARD) kernels [ 73 ]. The relevance of the features can be demonstrated with a polar plot according to Ayala Izurieta et al [ 74 ]. Hereby, the class-specific relevance for each feature is obtained, with values further away from the center expressing increasing importance.…”

Section: Resultsmentioning

confidence: 99%

“…Furthermore, summary percentages per class are provided. Right : polar plot of GPC band relevance for the four classes calculated according to the equations described in [ 74 ]. The further away from center, the more important.…”

Section: Figurementioning

confidence: 99%

Introducing ARTMO’s Machine-Learning Classification Algorithms Toolbox: Application to Plant-Type Detection in a Semi-Steppe Iranian Landscape

et al. 2022

Self Cite

View full text Add to dashboard Cite

Accurate plant-type (PT) detection forms an important basis for sustainable land management maintaining biodiversity and ecosystem services. In this sense, Sentinel-2 satellite images of the Copernicus program offer spatial, spectral, temporal, and radiometric characteristics with great potential for mapping and monitoring PTs. In addition, the selection of a best-performing algorithm needs to be considered for obtaining PT classification as accurate as possible . To date, no freely downloadable toolbox exists that brings the diversity of the latest supervised machine-learning classification algorithms (MLCAs) together into a single intuitive user-friendly graphical user interface (GUI). To fill this gap and to facilitate and automate the usage of MLCAs, here we present a novel GUI software package that allows systematically training, validating, and applying pixel-based MLCA models to remote sensing imagery. The so-called MLCA toolbox has been integrated within ARTMO’s software framework developed in Matlab which implements most of the state-of-the-art methods in the machine learning community. To demonstrate its utility, we chose a heterogeneous case study scene, a landscape in Southwest Iran to map PTs. In this area, four main PTs were identified, consisting of shrub land, grass land, semi-shrub land, and shrub land–grass land vegetation. Having developed 21 MLCAs using the same training and validation, datasets led to varying accuracy results. Gaussian process classifier (GPC) was validated as the top-performing classifier, with an overall accuracy (OA) of 90%. GPC follows a Laplace approximation to the Gaussian likelihood under the supervised classification framework, emerging as a very competitive alternative to common MLCAs. Random forests resulted in the second-best performance with an OA of 86%. Two other types of ensemble-learning algorithms, i.e., tree-ensemble learning (bagging) and decision tree (with error-correcting output codes), yielded an OA of 83% and 82%, respectively. Following, thirteen classifiers reported OA between 70% and 80%, and the remaining four classifiers reported an OA below 70%. We conclude that GPC substantially outperformed all classifiers, and thus, provides enormous potential for the classification of a diversity of land-cover types. In addition, its probabilistic formulation provides valuable band ranking information, as well as associated predictive variance at a pixel level. Nevertheless, as these are supervised (data-driven) classifiers, performances depend on the entered training data, meaning that an assessment of all MLCAs is crucial for any application. Our analysis demonstrated the efficacy of ARTMO’s MLCA toolbox for an automated evaluation of the classifiers and subsequent thematic mapping.

show abstract

“…In a study by Emadi et al (2020) using support vector machine (SVM), artificial neural network (ANN), random forest (RF), deep neural network and extreme gradient boost, SOC contents ranged from 0.02% to 11.48% with a mean and standard deviation of 2.19% and 1.27% respectively. Ayala Izurieta et al (2022) revealed that for the 0-30cm depth the SOC values ranged between 3.8-16.4%. As a result, in the 0-30cm depth, the SOC % prediction model had an R 2 of 0.85, RMSE of 1.58% and the SOC (Mgha -1 ) reached, R 2 = 0.79, RMSE = 24.7.…”

Section: Introductionmentioning

confidence: 98%

Estimating Soil Organic Carbon using Sentinel-2 data under Zero Tillage Agriculture: A Machine Learning Approach

Mango,

Narissara,

Som-ard

2024

Preprint

View full text Add to dashboard Cite

Soil organic carbon (SOC) is the main component of soil organic matter (SOM) and constitutes the crucial ingredient of the soil. It supports key soil functions, stabilizes soil structure, aid in plant nutrients retention and release, and promote water infiltration and storage. Predicting SOC using Sentinel-2 images integrated with machine learning algorithms under zero tillage practice is inadequately documented for developing countries like Zimbabwe. The purpose of this study is to evaluate the integrated use of Sentinel-2 data together with machine learning (ML) algorithms for SOC estimation. The support vector machine (SVM), artificial neural network (ANN), and partial least square regression (PLSR) models were used with a cross-validation to estimate the SOC content based on 50 georeferenced calibration samples under a zero-tillage practice. The ANN model outperformed the other two models by delivering a coefficient of determination (R2) of between 55-60% of SOC variability and RMSE varies between 5.01-8.78%, whereas for the SVM, R2 varies between 0.53-0.57 and RMSE varies between 6.25-11.39%. The least estimates of SOC are provided by the PLSR algorithm (R2 = 0.44–0.49 and RMSE = 7.59-12.42% for the top 15cm depth SOC estimation. Results with RMSE, MAE and R2 for SVM, ANN and PLSR, show that the ANN model is able to capture SOC variability. Although the ANN algorithm provides more accurate SOC estimates than the SVM algorithm, the difference in accuracy is not significant. Results revealed a satisfactory agreement between the SOC and zero tillage practice (R2, coefficient of variation (CV), mean absolute error (MAE), and root mean square error (RMSE) using SVM, ANN and PLSR for the validation dataset using four predictor variables. The calibration results of SOC indicated that the mean SOC was 5.19% and the validation mean SOC was 5.02%. The calibration dataset had higher degree of variation (27.94%) around its mean compared to the validation dataset (24.30%). The accurate map results can be used as an important tool for informed decisions about soil health and productivity by the farmers, land managers and policy makers.

show abstract

Improving the remote estimation of soil organic carbon in complex ecosystems with Sentinel-2 and GIS using Gaussian processes regression

Cited by 23 publications

References 104 publications

Evaluation of Soil Organic Carbon Storage of Atillo in the Ecuadorian Andean Wetlands

Evaluation of Soil Organic Carbon Storage of Atillo in the Ecuadorian Andean Wetlands

Introducing ARTMO’s Machine-Learning Classification Algorithms Toolbox: Application to Plant-Type Detection in a Semi-Steppe Iranian Landscape

Estimating Soil Organic Carbon using Sentinel-2 data under Zero Tillage Agriculture: A Machine Learning Approach

Contact Info

Product

Resources

About