The Influence of Downscaling Models and Observations on Future Hydrochemistry Reponses of Forest Watersheds

Abstract:

Most projections of climate change impacts on ecosystems rely on multiple climate model projections, but utilize only one downscaling approach trained on one set of observations. Here, we explore the extent to which modeled biogeochemical responses to changing climate are affected by the selection of the climate downscaling method and training observations used in the complex mountainous terrain of the Hubbard Brook Experimental Forest, New Hampshire, USA. We evaluated three different downscaling methods: the monthly delta method (or the “change factor method”); monthly quantile mapping (Bias Correction-Spatial Disaggregation, or BCSD); and daily quantile regression (Asynchronous Regional Regression Model, or ARRM). Additionally, we trained outputs, from four AOGCMs (CCSM3, HadCM3, PCM, and GFDL-CM2) driven by higher (A1fi) and lower (B1) future emission scenarios, on two sets of observations (1/8^th degree resolution grid vs. individual weather station) to generate the high-resolution climate input for the hydrochemical model PnET-BGC (ensemble of 48 runs).

The choice of downscaling approach and spatial resolution of the observations used to train the downscaling model both had a major impact on modeled soil moisture and streamflow which in turn affected forest growth, net nitrification and stream chemistry. Specifically, the delta method, the simplest downscaling approach evaluated, was highly sensitive to the observations used, resulting in projections that were significantly different than those produced with the BCSD and ARRM methods. Using spatially smoothed gridded observations and/or methods that do not resolve sub-monthly shifts in the distribution of temperature and/or precipitation can produce poor results in model applications run at higher temporal and/or spatial resolutions. These results underscore the importance of carefully considering the observations and downscaling method used to generate climate change projections for smaller scale modeling studies.