Accuracy of regional-to-global soil maps for on-farm decision making: Are soil maps good enough?

Authors: MAYNARD, JONATHAN - University Of Colorado; YEBOAH, EDWARD - Csir-Crops Research Institute; OWUSU, STEVEN - Csir-Crops Research Institute; BUENEMANN, MICHAELA - New Mexico State University; NEFF, JASON - University Of Colorado; Herrick, Jeffrey - Jeff

Submitted to: Agriculture, Ecosystems and Environment
Publication Type: Peer Reviewed Journal
Publication Acceptance Date: 4/21/2022
Publication Date: 5/19/2022
Citation: Maynard, J.J., Yeboah, E., Owusu, S., Buenemann, M., Neff, J.C., Herrick, J.E. 2022. Accuracy of regional-to-global soil maps for on-farm decision making: Are soil maps good enough? Agriculture, Ecosystems and Environment. https://doi.org/10.5194/egusphere-2022-246.
DOI: https://doi.org/10.5194/egusphere-2022-246

Interpretive Summary: Soil maps are often accepted as "truth". The reality is that soil maps are simply predictions of the soils, and soil properties, occurring at a particular location. Most traditional soil maps are actually maps of groups of soils that occur together within map units, while the newer "digital" soil maps use statistical methods to predict soil properties for each pixel. The accuracy of the predictions, varies widely across the globe and, is poorly understood. Even less well understood is whether the accuracy is high enough to use the prediction to make a decision, without digging a hole to confirm soil type. . The main objective of this study was to evaluate the accuracy of four soil maps currently in use in Ghana using a collection of 6,959 soil profile descriptions recorded using the LandPKS mobile application. Results from this study revealed that current soil maps in Ghana, while often effective for larger-scale analyses, lack the needed accuracy to reliably inform soil management decisions at the farm-scale.

Technical Abstract: A major obstacle to selecting the most appropriate crops and closing the yield gap in many areas of the world is a lack of site-specific soil information. Accurate information on soil properties is critical for identifying soil limitations and the management practices needed to improve crop yields. However, acquiring accurate soil information is often difficult due to the high spatial and temporal variability of soil properties at small scales (i.e., landscape-to-farm-scale, sub-decadal-scale), and the cost and variable reliability of laboratory based chemical and physical analyses. It has been hoped that soil maps could someday fill this need, and with recent advancements in predictive soil mapping there is a growing expectation that soil map predictions can provide much of the information needed to inform soil management. The main objective of this study was to evaluate the accuracy of four soil maps currently in use in Ghana using a collection of 6,959 soil profile descriptions recorded using the LandPKS mobile application. Results from this study revealed that current soil maps in Ghana, while often effective for larger-scale analyses, lack the needed accuracy to reliably inform soil management decisions at the farm-scale. Standard measures of map accuracy for soil texture class and rock fragment class showed that all soil maps were equally inaccurate in estimating the correct property class, with overall accuracies ranging from 7-11% for soil texture classes and 27-29% for soil rock fragment classes. When accounting for class-adjacency (i.e., a ‘correct’ prediction includes classes adjacent in property space to the measured class) overall accuracies increased to 36-41% for soil texture and 43-49% for soil rock fragments. This indicates that 50-60% of map-based soil property estimates were significantly different (i.e., greater than one property class difference) than site-based estimates. To better understand the functional implications of these soil property differences, we used a modified version of the Global Agro-Ecological Zone soil suitability modeling framework to derive soil suitability ratings for each soil data source. Using a low-input, rain-fed, maize production scenario, we evaluated the functional accuracy of map-based soil property estimates and found that all map-based suitability ratings were classified as having slight to no constraints, in contrast to the site-based suitability ratings where over 70% of sites were classifieds as having moderate-to-severe soil constraints. In this study area soil map data significantly overestimated crop suitability for most study sites, potentially leading to ineffective agronomic investments by cash-constrained smallholder farmers.