There is a new paper which presents forecasts of regional climate decades into the future. Climate Science has been very critical of such papers since i) we have wait decades to validate the skill of the predictions, and ii) the multi-decadal global climate models have shown no skill in predicting regional climate events (such as drought) in the 20th century up to the present (e.g. see
However, in contrast to many other such papers, this paper does recognize the serious limitations of this approach.
The paper is
Harmsen, E.W., Miller, N.L., Schlegel, N.J., Gonzalez, J.E., 2009: Seasonal climate change impacts on evapotranspiration, precipitation deficit and crop yield in Puerto Rico. Agricultural Water Management, 96, issue 7, 1085 – 1095,
with the abstract
“The purpose of this study was to estimate precipitation (P), reference evapotranspiration (ETo), precipitation deficit (PD = P − ETo) and relative crop yield reduction (YR) for a generic crop under climate change conditions for three locations in Puerto Rico: Adjuntas, Mayaguëz, and Lajas. Reference evapotranspiration was estimated by the Penman-Monteith method. Precipitation and temperature data were statistically downscaled and evaluated using the DOE/NCAR PCM global circulation model projections for the B1 (low), A2 (mid-high) and A1fi (high) emission scenarios of the Intergovernmental Panel on Climate Change Special Report on Emission Scenarios. Relative crop yield reduction was estimated from a water stress factor, which is a function of soil moisture content. Average soil moisture content for the three locations was determined by means of a simple water balance approach.
Results from the analysis indicate that the rainy season will become wetter and the dry season will become drier. The 20-year average September precipitation excess (i.e., PD > 0) increased for all scenarios and locations from 121 to 321 mm between 2000 and 2090. Conversely, the 20-year average February precipitation deficit (i.e., PD < 0) changed from —27 to —77 mm between 2000 and 2090. The results suggest that additional water could be saved during the wet months to offset increased irrigation requirements during the dry months. The 20-year average relative crop yield reduction for all scenarios decreased on average from 12% to 6% between 2000 and 2090 during September, but increased on average from 51% to 64% during February. Information related to the components of the hydrologic water budget(i.e., actual evapotranspiration, surface runoff, aquifer recharge and soil moisture storage) is also presented. This study provides important information that may be useful for future water resource planning in Puerto Rico.”
The cavaets to this study are given in the section “Limitations in results presented”, where they write
“The results presented in this paper should necessarily be viewed with caution since they are based in part on coarse resolution GCM data downscaled to single sites. As Pielke et al. (2007) rightly point out, future ‘‘agricultural impacts extend far beyond a global mean temperature and include other anthropogenic climate forcings.’’ Some of these forcings include land-use change, atmospheric aerosols, and complex nonlinear feedbacks, not accounted for in present-day, and likely next-generation, GCMs. Statistical downscaling itself assumes that the predictor– predictand relationship remains constant in time with stationary dynamic conditions under future climate change (Mearns et al., 2003). Furthermore, this study was based on only one GCM and since many uncertainties still exist among different models, the results need to be used with caution (Bouraoui et al., 1997).”
A section in every paper of this type titled “Limitations in results presented”, should be a requirement so that readers can more appropriately place such studies in ther proper perspective as “sensitivity studies”.