The Land Potential Knowledge System: Utilizing Crowdsourced Data and Cloud-Based Analytics to Predict Soil Productivity and Degradation Risk.

Globally, the productivity of smallholder cropping and grazing systems is constrained by soil properties and degradation processes. Soils are managed at the field scale, but soil maps are generally too coarse to provide accurate soil property values necessary to estimate productivity and erosion potential, and to support management recommendations. The problem is particularly acute in the developing world, where soil information is extremely limited. The Land Potential Knowledge System (LandPKS) is currently being developed to crowdsource site-specific soil and landscape data as a means of generating localized estimates of productivity and degradation risk, and to deliver these estimates to users in near-real time. The system consists of a modular suite of mobile phone applications and cloud-based analytics driven by regional weather, soil and topographic data. The core application guides the user through a site characterization process that includes land use, topography and soil profile characteristics. User data are utilized as inputs in the APEX model to estimate productivity and erosion resistance of the desired land management at the site. Model outputs are then used to create relative indices of productivity and degradation risk for specific sites. In the future, the system will provide users with management options and connect land managers with similar land potential. It will also leverage the widespread use of mobile phones to crowdsource information, including follow-up monitoring data, in data poor regions for improving model predictions. The system is being piloted in 2014 in Kenya and Namibia, with an initial focus on drylands. For current information, please see www.landpotential.org.