cc: "Linda Mearns" date: 13 May 1999 06:25:47 -0600 from: "Tim Kittel" subject: IPCC comments to: "Mike Hulme" REGARDING IPCC comments Mike: Thanks for the opportunity to contribute to your chapter with Linda on climate change scenario development. Apologies for not providing material sooner. Here are a few comments and suggestions. I'll send remaining cites in a subsequent email. Best, Tim p.s.* After 13 May 99, my 'qgate.ucar.edu' address will not be valid. Please reply to: kittel@ucar.edu -------- (1) Section 13.3.1, parag. 1-- "Kittel/VEMAP" References for VEMAP2 scenario development are: Kittel et al. 1997, and in prep.(b) (2) Section 13.3.1, parag. 2 -- Comment on last sentence: An additional advantage to using a more recent baseline period (e.g., 1961 1990) is that selecting a period near the end of the historical record reduces the likelihood of there being a strong step change at the transition from the historical series to the model-simulated future series. This is particularly important if a GCM has not adequately captured regional historical trends from an earlier baseline period (e.g., 1931-60) to the end of the simulated historical record. Such a divergence between observed and simulated historical trends means that adjustment of the GCM series to the earlier baseline period will not remove all model biases present at the transition point. Causes of divergent traces include omission of key historical forcings (e.g., sulfate aerosols, solar variability) in GCM runs. (3) Section 13.3.2, 1st parag and material outlined in brackets -- Additional material re sources of differences and uncertainty in observations (though to some extent this is overlapping with your 2nd sentence, first paragraph): A comparison of global climatologies used to evaluate regional biases in GCMs in previous IPCC assessment reports (IPCC 1990, 1996) showed considerable differences at the regional level among datasets (Kittel et al. 1998). Specifically, the coefficient of variation in estimates from four datasets of regional mean winter and summer precipitation ranged from 5% to over 40% and the standard deviation of dataset estimates of regional mean temperatures ranged up to ~1 deg C. Some of this uncertainty in observed climatology likely arose from differences in station density and interpolation methods, as well as from different averaging periods. Selection of interpolation technique is an important factor in reducing uncertainty in regional datasets especially where station density is low relative to regional heterogeneity in controls over climate. A common problem that some techniques endeavor to correct is systematic biases in station locations (e.g., toward low elevation sites). Among alternative techniques currently used to spatially interpolate station records (Daly et al. 1994, 1999, Hutchinson 19xx, Cramer et al. 199x, Willmott et al. [ref.?], New 1999?, Kittel et al., in prep (a)), some require and can take advantage of more information that others when station densities are higher while others likely perform better when station densities are low. However, this has not been systematically evaluated across a wide range of conditions. Gauge correction for precipitation data is another important factor in reducing uncertainty in baseline climatologies (Legates and Willmott 1990). However, this is problematic for many regions where metadata (station histories) and ancillary data (e.g., wind speed) are inadequate or where understanding of controls over gauge biases is poorly known (Groisman ref?). (4) Section 13.3.3, parag. 1, sentence 2 -- "Inadequate ... resolution, or else inaccurate model-simulated climates ..." A suggestion to clarify this point: Impact studies rarely use GCM outputs directly because GCM biases are too great and spatial resolution too coarse to get realistic assessment model simulations. (5) Section 13.3.3, last parag. -- differences or ratios. I suggest adding sentences giving when and why ratios are commonly used over diff's and vice versa: Differences are commonly applied to temperature, while ratios to those variables such as precipitation, vapor pressure, and radiation that are either positive or can easily reach (or approach) zero under surface conditions. The advantage of using ratios for these variables is that if, for example, the GCM bias for precipitation is very large over a desert then proportionally small changes in modeled precipitation can result in unrealistically large absolute changes, with large negative changes giving less than zero precipitation when applied to the baseline. In this situation, it is more straightforward and realistic to apply ratio changes. Ratios are often capped (e.g., at 5.0) so as to put reasonable bounds on the absolute magnitude of potential changes (Jenne 1992). (6) Cites for references given above -- Cramer et al 199x, - new PIK climate ref Daly, C., R.P. Neilson, and D.L. Phillips. 1994. A statistical-topographic model for mapping climatological precipitation over mountainous terrain. J. Appl. Meteorol. 33:140-158. Daly et al. 1999. Groisman, P. -Gauge correction problems ref New 1999 Hutchinson 19xx, Jenne, R.L. (1992) Climate model description and impact on terrestrial climate. Pages 145-164 in: Global climate change: implications, challenges and mitigation measures. S.K. Majumdar, L.S. Kalkstein, B. Yarnal, E.W. Miller, and L.M. Rosenfeld (eds), p. 566. Pennsylvania Academy of Science. Kittel, T.G.F., J.A. Royle, C. Daly, N.A. Rosenbloom, W.P. Gibson, H.H. Fisher, D.S. Schimel, L.M. Berliner, and VEMAP2 Participants. 1997. A gridded historical (1895-1993) bioclimate dataset for the conterminous United States. Pages 219-222, in: Proceedings of the 10th Conference on Applied Climatology, 20-24 October 1997, Reno, NV. American Meteorological Society, Boston. Kittel, T.G.F., J.A. Royle, C. Daly, N.A. Rosenbloom, P. Thornton, W.P. Gibson, H.H. Fisher, D.S. Schimel, L.M. Berliner, and VEMAP2 Participants. The VEMAP Phase 2 bioclimate database. I: A gridded historical (1895 1993) climate dataset for modeling ecosystem dynamics across the United States. Climate Dynamics, in preparation (a). Kittel, T.G.F., N.A. Rosenbloom, H.H. Fisher, D.S. Schimel, J.A. Royle, and VEMAP2 Participants. The VEMAP Phase 2 bioclimate database. II: Transient climate change scenarios for modeling 21st-century ecosystem dynamics across the United States. Climate Dynamics, in preparation (b). Legates DR,Willmott CJ (1990b) Mean seasonal and spatial variability in gauge-corrected, global precipitation. Intl J Climatol 10:111-127 Willmott et al. [interpolation scheme ref] ---