GC13G-0743:
The Impact of Changing Snowmelt Timing on Non-Irrigated Crop Yield: A Parametric and Non-Parametric Approach

Monday, 15 December 2014
Erin M Murray1, Kelly Cobourn2, Alejandro N Flores1 and Jennifer L Pierce1, (1)Boise State University, Department of Geosciences, Boise, ID, United States, (2)Virginia Polytechnic Institute and State University, Department of Forest Resources and Environmental Conservation, Blacksburg, VA, United States
Abstract:
As climate changes, the final date of spring snowmelt is projected to occur earlier in the year within the western United States. This earlier snowmelt timing may impact crop yield in snow-dominated watersheds by changing the timing of water delivery to agricultural fields. There is considerable uncertainty about how agricultural impacts of snowmelt timing may vary by region, crop-type, and practices like irrigation vs. dryland farming. Establishing the relationship between snowmelt timing and agricultural yield is important for understanding how changes in large-scale climatic indices (like snowmelt date) may be associated with changes in agricultural yield. A better understanding of the influence of changes in snowmelt on non-irrigated crop yield may additionally be extrapolated to better understand how climate change may alter biomass production in non-managed ecosystems. We utilized parametric regression techniques to isolate the magnitude of impact snowmelt timing has had on historical crop yield independently of climate and spatial variables that also impact yield. To do this, we examined the historical relationship between snowmelt timing and non-irrigated wheat and barley yield using a multiple linear regression model to predict yield in several Idaho counties as a function of snowmelt date, climate variables (precipitation and growing degree-days), and spatial differences between counties. We utilized non-parametric techniques to determine where snowmelt timing has positively versus negatively impacted yield. To do this, we developed classification and regression trees to identify spatial controls (e.g. latitude, elevation) on the relationship between snowmelt timing and yield.

Most trends suggest a decrease in crop yield with earlier snowmelt, but a significant opposite relationship is observed in some regions of Idaho. An earlier snowmelt date occurring at high latitudes corresponds with higher than average wheat yield. Therefore, Northern Idaho may benefit from planting wheat in years with early snowmelt timing. Barley grown in counties with low elevation and low topography may also benefit from early snowmelt timing. Despite the beneficial relationship in some counties, wheat and barley yield have been lower on average in past years with early snowmelt timing in most of Idaho.