Modeling droughty soils at regional scales in Pacific Northwest Forests, USA

• Published in: Forest ecology and management Vol. 424; pp. 121 - 135
• Main Authors: Ringo, Chris, Bennett, Karen, Noller, Jay, Jiang, Duo, Moore, David
• Format: Journal Article
• Language: English
• Published: Elsevier B.V 15.09.2018
• Subjects: Evapotranspiration; Landscape model; Logistic regression; Soil moisture deficit; Soil moisture drought; Logistic regression; Landscape model; Soil moisture drought; Evapotranspiration; Soil moisture deficit
• ISSN: 0378-1127; 1872-7042
• DOI: 10.1016/j.foreco.2018.04.019

•We develop a new approach to modeling potentially drought-prone soils.•We develop a working definition of a “droughty soil” from SNOTEL soil moisture data.•Broad-scale climatic data adds value to soil moisture storage capacity estimates.•Site-specific soil moisture data can be usefully extrapolated to landscape scales. Natural resource managers need better estimates of water storage and supply in forested landscapes. These estimates would aid planning for management activities that maintain and enhance forest health and productivity and help prepare forested landscapes for a changing climate. In particular, low soil moisture in combination with high evaporative demands can induce significant stresses on forests, increasing vulnerability to attacks of insect and disease, as well as increasing wildfire risk. Although high-resolution soils data exist for much of the Pacific Northwest, regional-scale datasets that identify forested areas potentially vulnerable to soil moisture-related drought do not exist. In this study, we used readily available spatial datasets depicting available water supply, soil depth, and evapotranspiration to model the likelihood that soils experience prolonged summer drying. To calibrate the model, we examined soil profile descriptions, lab data, and soil moisture curves for 25 sites throughout the Pacific Northwest and estimated the average annual number of days that soil moisture drops to levels at or below the permanent wilting point, a theoretical lower limit of plant-available water. Using this approach, we found statistically significant relationships between the independent variables and broad classes of soil moisture levels representing the highest and lowest levels of plant-available moisture. We then used these relationships to create a landscape-level droughty soil index for the Pacific Northwest. We expect that this approach can be further developed to include additional soil moisture data outside Washington and Oregon and enhanced with other explanatory variables such as topographic position, elevation, and vegetation type. With the addition of vegetation-related data, in particular, the current modeling approach can aid in identifying vulnerable landscapes in the context of managing for increased forest resiliency in the Pacific Northwest.