Abstract:Increasingly, applications of machine learning techniques for digital soil mapping (DSM) are being used for different soil mapping purposes. Considering the variety of models available, it is important to know their performance in relation to soil data and environmental variables involved in soil mapping. This paper investigated the performance of eight machine learning algorithms for soil mapping in a tropical mountainous area of an official rural settlement in the Zona da Mata region in Brazil. Morphometric … Show more
“…Different techniques of DSM have been widely assessed in studies in tropical soils 23–27 . However, mountainous and complex relief areas still present challenges to any approach to soil mapping, due to the complex and scale-dependent interactions among soil forming factors and the cost-effort associated to survey inaccessible areas.…”
This study aimed to evaluate the performance of three spatial association models used in digital soil mapping and the effects of additional point sampling in a steep-slope watershed (1,200 ha). A soil survey was carried out and 74 soil profiles were analyzed. The tested models were: Multinomial logistic regression (MLR), C5 decision tree (C5-DT) and Random forest (RF). In order to reduce the effects of an imbalanced dataset on the accuracy of the tested models, additional sampling retrieved by photointerpretation was necessary. Accuracy assessment was based on aggregated data from a proportional 5-fold cross-validation procedure. Extrapolation assessment was based on the multivariate environmental similarity surface (MESS). The RF model including additional sampling (RF*) showed the best performance among the tested models (overall accuracy = 49%, kappa index = 0.33). The RF* allowed to link soil mapping units (SMU) and, in the case of less-common soil classes in the watershed, to set specific conditions of occurrence on the space of terrain-attributes. MESS analysis showed reliable outputs for 82.5% of the watershed. SMU distribution across the watershed was: Typic Rhodudult (56%), Typic Hapludult* (13%), Typic Dystrudept (10%), Typic Endoaquent + Fluventic Dystrudept (10%), Typic Hapludult (9.5%) and Rhodic Hapludox + Typic Hapludox (2%).
“…Different techniques of DSM have been widely assessed in studies in tropical soils 23–27 . However, mountainous and complex relief areas still present challenges to any approach to soil mapping, due to the complex and scale-dependent interactions among soil forming factors and the cost-effort associated to survey inaccessible areas.…”
This study aimed to evaluate the performance of three spatial association models used in digital soil mapping and the effects of additional point sampling in a steep-slope watershed (1,200 ha). A soil survey was carried out and 74 soil profiles were analyzed. The tested models were: Multinomial logistic regression (MLR), C5 decision tree (C5-DT) and Random forest (RF). In order to reduce the effects of an imbalanced dataset on the accuracy of the tested models, additional sampling retrieved by photointerpretation was necessary. Accuracy assessment was based on aggregated data from a proportional 5-fold cross-validation procedure. Extrapolation assessment was based on the multivariate environmental similarity surface (MESS). The RF model including additional sampling (RF*) showed the best performance among the tested models (overall accuracy = 49%, kappa index = 0.33). The RF* allowed to link soil mapping units (SMU) and, in the case of less-common soil classes in the watershed, to set specific conditions of occurrence on the space of terrain-attributes. MESS analysis showed reliable outputs for 82.5% of the watershed. SMU distribution across the watershed was: Typic Rhodudult (56%), Typic Hapludult* (13%), Typic Dystrudept (10%), Typic Endoaquent + Fluventic Dystrudept (10%), Typic Hapludult (9.5%) and Rhodic Hapludox + Typic Hapludox (2%).
“…We preferred generating a silt content layer because a future objective is to transfer these models to other regions in Kenya that have more silty than sandy and clayey soils. We used co-kriging in ArcMap to interpolate the field estimated silt content together with the Modified Soil-Adjusted Vegetation Index (MSAVI2), TWI, and slope as covariates, since previous literature had mentioned their importance in representing soil texture variability [ 28 – 30 ]. Fig.…”
Background
African trypanosomiasis, which is mainly transmitted by tsetse flies (Glossina spp.), is a threat to public health and a significant hindrance to animal production. Tools that can reduce tsetse densities and interrupt disease transmission exist, but their large-scale deployment is limited by high implementation costs. This is in part limited by the absence of knowledge of breeding sites and dispersal data, and tools that can predict these in the absence of ground-truthing.
Methods
In Kenya, tsetse collections were carried out in 261 randomized points within Shimba Hills National Reserve (SHNR) and villages up to 5 km from the reserve boundary between 2017 and 2019. Considering their limited dispersal rate, we used in situ observations of newly emerged flies that had not had a blood meal (teneral) as a proxy for active breeding locations. We fitted commonly used species distribution models linking teneral and non-teneral tsetse presence with satellite-derived vegetation cover type fractions, greenness, temperature, and soil texture and moisture indices separately for the wet and dry season. Model performance was assessed with area under curve (AUC) statistics, while the maximum sum of sensitivity and specificity was used to classify suitable breeding or foraging sites.
Results
Glossina pallidipes flies were caught in 47% of the 261 traps, with teneral flies accounting for 37% of these traps. Fitted models were more accurate for the teneral flies (AUC = 0.83) as compared to the non-teneral (AUC = 0.73). The probability of teneral fly occurrence increased with woodland fractions but decreased with cropland fractions. During the wet season, the likelihood of teneral flies occurring decreased as silt content increased. Adult tsetse flies were less likely to be trapped in areas with average land surface temperatures below 24 °C. The models predicted that 63% of the potential tsetse breeding area was within the SHNR, but also indicated potential breeding pockets outside the reserve.
Conclusion
Modelling tsetse occurrence data disaggregated by life stages with time series of satellite-derived variables enabled the spatial characterization of potential breeding and foraging sites for G. pallidipes. Our models provide insight into tsetse bionomics and aid in characterising tsetse infestations and thus prioritizing control areas.
Graphical abstract
“…For CCE, EC, and pH, remote sensing data was particularly effective due to the accumulation of salts at the surface of the soil, which was easily detected by RS imagery. Meier et al [67] selected 10 covariates for soil mapping, including four topographic covariates, three images from Landsat, two climatic maps, and the map of Euclidean distance from the drainage network. This study showed that MRVBF, temperature, rainfall, and TWI were the most important covariates for soil mapping (Figure 3).…”
This study tested and evaluated a suite of nine individual base learners and seven model averaging techniques for predicting the spatial distribution of soil properties in central Iran. Based on the nested-cross validation approach, the results showed that the artificial neural network and Random Forest base learners were the most effective in predicting soil organic matter and electrical conductivity, respectively. However, all seven model averaging techniques performed better than the base learners. For example, the Granger–Ramanathan averaging approach resulted in the highest prediction accuracy for soil organic matter, while the Bayesian model averaging approach was most effective in predicting sand content. These results indicate that the model averaging approaches could improve the predictive accuracy for soil properties. The resulting maps, produced at a 30 m spatial resolution, can be used as valuable baseline information for managing environmental resources more effectively.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
