Digital mapping of soil ecosystem services in Scotland using neural networks and relationship modelling—Part 1: Mapping of soil classes

Abstract

A digital mapping approach was applied to soils in Scotland, producing maps at 100-m resolution and different levels of classification. This used neural networks to predict fuzzy soil class weightings based upon site descriptors from existing soil survey data. The intention of this work was to produce a set of soil maps for Scotland, which provide greater spatial resolution mapping than currently available, and provide fuzzy data to be used in mapping of ecosystem services using a novel approach explained in a second paper. When selecting the class with the highest weighting for test data points, Producer's Accuracy of 70.7% was achieved in mapping the broadest classification (five classes), while Producer's Accuracy levels of 59.9% and 43.7%were achieved for 11 and 30 classes, respectively. Evaluation of confusion matrices generated from the neural network model tests showed that, particularly for the classification systems with 11 and 30 classes, misclassification errors were more common between similar soil classes than between classes that were very different from one another. This implies that straightforward estimation of classification accuracy that measures "correct" and "incorrect" gives misleading results, and that the fuzzy classification maps are more useful than the crisp classification accuracy values.These fuzzy classification maps are therefore potentially useful for evaluating ecosystem services relating to soil type. We conclude that the mapping approach used here provides a classification accuracy comparable to other approaches and that the outputs can be used for mapping soil properties, processes, functions and ecosystem services. K E Y W O R D Sdigital soil mapping, land management, remote sensing, soil classification, spatial resolution 206 |

Digital mapping of soil texture classes for efficient land management in the Piedmont plain of Iran

“…As presented in Table 6, the user's accuracy for the clay loam class for the training dataset was found to be higher than the producer's accuracy. This indicates that the clay loam class as shown on the map is accurate, but their actual presence on the ground is less likely to be detected (Aitkenhead & Coull, 2019). In other words, the existence of the clay loam class is underestimated by the model.…”

“…In Table 6, the producer's accuracy for the loam class in the training set was found to be higher than the user's accuracy. This indicates that the loam class tends to be more accurately identified in the field, but their presence on the map is probably incorrect, an indication of overestimation of the loam class by the model (Aitkenhead & Coull, 2019). This class is more than half of the observed dataset, so this result is expected in data‐driven modelling methods (Wadoux et al, 2021).…”

Delineation of high‐resolution soil carbon management zones using digital soil mapping: A step towards mitigating climate change in the Northeastern Himalayas, India

“…However, the availability of dense sampling intensity along with precise spatial mapping determines the success of surface maps (Yao et al., 2013). The effect of environmental covariates such as climate patterns, parent material, topography and soil management vary to a great extent in a spatio‐temporal framework of nature (Aitkenhead & Coull, 2019; Bian et al., 2020; Dharumarajan et al., 2021; Dharumarajan & Hegde, 2020; Duarte et al., 2022; Li, Liu, et al., 2022; Li, Yue, et al., 2022; Mallik et al., 2022; Wang et al., 2022). Therefore, an empirical modelling based on mathematical and statistical methods like digital soil mapping (DSM) that takes into account anthropogenic and environmental factors for the prediction of SOC is more realistic than the geostatistical interpolation techniques (Dharumarajan et al., 2021; Reddy et al., 2021).…”