The project reported here examined sensitivities of prairie soil landscapes to climate change, variability and extreme hydroclimatic events. Early in the project, a review of existing climate impact assessments and methodologies suggested that most are based on an incomplete understanding of the climate forcing of geomorphic systems, especially in relation to the influence of scale on the understanding and modeling of biophysical systems. Therefore a high priority was given to developing a practical framework for the assessment of potential impacts of climate change on the soil landscapes of the Canadian plains. This framework 1) facilitates transfer of the results of scientific research to stakeholders for the planning of adaptation to soil landscape sensitivity, and 2) is spatially-explicit, unlike models and methods that lack geographic reference and thus direct application to the real world. This approach to climate impact assessment requires a spatial data model and a model of the geomorphic response to climate. Since landforms are the product of geomorphic processes, assemblages of landforms and the associated soil profiles, are the geographic expression of a geomorphic systems. The use the soil landscape as a spatial data structures expands climate impact assessment beyond the study of soil loss or and beyond the farm or field. Conventional approaches to the assessment of soil erosion risk, estimating potential soil loss from fields, will not support the integrated planning over large areas or thus the adaptation to the impacts of climate change.
The response of geomorphic systems to climate is complex. Long periods of landscape stability are interrupted by short bursts of erosion as s system responds to the forces of a hydroclimatic event that exceeds a geomorphic threshold. Irreversible landscape change can occur in response to single events. Much of what we know about the climatic forcing of natural systems is base on detailed field experiments conducted over small areas in contrast to the scale of land and water management. A “scaling down” of climate and a “scaling up” of process data is required for the study of climate impacts on soil landscapes. Most process simulation models fail to work when scaled up because of the greater complexity of larger systems and non-linearity caused by feedback among system variables, and the emergence of characteristic patterns and processes at coarser scales. Virtually all existing models of soil loss and landscape change are inappropriate for the spatial analysis of the climatic forcing of surface processes in the Canadian plains. The greatest promise for a model-based approach lies in relatively simple physically-based models, because they are more scale robust than empirical models (i.e. derived from the measurement of erosion from plots) or the data-rich analytical modelling of slope and channel processes. Models reduced to a form that capture essential or salient factors are most easily applied to assessment of soil landscape sensitivity at a regional. In fact dimensionless indices are simple and practical, yet meaningful, models of landscape sensitivity. Mapping the Aridity Index (Precipitation/Potential Evapotranspiration) for 1961-90 and the 2050s demonstrates the the area of land at risk of desertification will increase by about 50%.
In managed landscapes erosion is mostly a socio-economic issue since erosion can be prevented by soil conservation, but capability and willingness to implement soil conservations are governed by a host of social and economic factors. Even though rates of erosion are managed, land managers must realize that landscape change is a threshold process, such that the conditions that lead to land degradation are established before they are recognized. An increase in theprobability of extreme erosion evens, as the result of climate change, above “once in a lifetime”may justify increased use of soil conservation.
