Regional-scale assessment of soil functions and resilience indicators: Accounting for change of support to estimate primary soil properties and their uncertainty.

Researchers and stakeholders are increasingly interested in evaluating soil functions and their relationships with soil physical resilience to prevent and mitigate natural and anthropogenic threats. This study aims to assess the effect of changing spatial support on estimating primary soil properties that are then used to map compound indicators of soil functions. The specific objectives are to determine the epistemic uncertainty associated with the estimates and assess the impact of accounting for the change-of-support when scaling up from the point- to the block-resolution. The study area is the southern Italian region of Campania, which is renowned for its high-quality agricultural and dairy products but is increasingly subject to soil disturbance and degradation. Seven primary soil properties (sand and clay contents, oven-dry soil bulk density, calcium carbonate, rock fragment, soil organic matter, and pH) were estimated over 75-m75-m blocks by jointly analyzing more than 3,000 point-based soil measurements with gridded terrain attributes by using multi-collocated block cokriging. Two compound soil indicators were then assessed from the maps of these soil properties: 1) the soil organic carbon stock (SOCS) in the entire region of Campania, and 2) the recharge transit time in the Sele alluvial plain where information about the mean annual depth to groundwater is available. The change-of-support is proved to affect the total variance and the assessment of the extremes in the probability distributions of the primary soil variables. Our results emphasize that negative consequences occur when the change-of-support problem, upscaling in particular, is overlooked as happens when analyzing environmental data, especially for planning purposes and in the case of environmental risk assessment