of Hans von Storch
The term "downscaling" was first used in the MPI Report #64 - Hans von Storch, Eduardo Zorita, Ulrich Cubasch (1991): Downscaling of Global Climate Change Estimates to Regional Scales: An Application to Iberian Rainfall in Wintertime. This approach was developped out of a request for elaborating the utility of GCM output, asked by a group of German hydrologists - the details of the meeting are by now forgotten. In this paper, Hans von Storch and Eduardo Zorita applied Canonical Correlation Analysis to derive Iberian Peninsula seasonal rainfall change in winter from global GCM scenarios
The work was inspired by Kim, J.W., J.-T. Chang, N.L. Baker, D.S. Wilks and W.L. Gates, 1984: The statistical problem of climate inversion: Determination of the relationship between local and large-scale climate. Mon. Wea. Rev. 112: 2069-2077, .
A first sketch was presented in
von Storch, H. and E. Zorita, 1990: Assessment of regional climate changes with the help of global GCM's: an example. CAS/JSC Working group on Numerical Experimentation. WMO Report no. 14, 7.30
Eventually the method and the analysis were finally published first in Spain and then in the Journal of Climate:
Zorita, E. and H.von Storch, 1991: Estimacion de cambios en la precipitacion en la Peninsula Iberica a partir de experimentos con modelos de circulacion general. In "Modelos en Meteorologia y Climatologia", Universidad Complutense, Madrid
von Storch, H., E. Zorita and U. Cubasch, 1993: Downscaling of global climate change estimates to regional scales: An application to Iberian rainfall in wintertime. - J. Climate 6: 1161-1171 - the MPI report #64 mentioned above was the preprint of this article.
In 2019, Hans von Storch and Eduardo Zorita published a kind of review of the foundations of empirical downscaling, and its roots in synoptic climatology and spatial interpolation:
von Storch, H., and E. Zorita, 2019: The history of ideas of empirical downscaling – from synoptic dynamics and spatial interpolation. Frontier (doi: 10.3389/fenvs.2019.00021; 8 pp.)
The problem of simulating the regional climate conditional upon the large scale state of a re-analysis or a GCM simulation is no longer formulated as a boundary-value problem but as a data-driven problem, with the known large-scale conditioning the medium- and small scale dynamics. In doing so, it becomes also clear that the expected value added is not in the large scales, which are already described by the global model/analysis but in those scales, which are insufficiently resolved by the global analysis/model.
von Storch, H., H. Langenberg and F. Feser, 2000: A spectral nudging technique for dynamical downscaling purposes. Mon. Wea. Rev. 128: 3664-3673.
Preliminary "grey" versions were
Langenberg, H, and H. von Storch, 1999: Spectral analysis in regionalisation retaining the large scales but adding detail. In H. Ritchie (Ed.): Research Activities in Atmospheric and Oceanic Modelling, WMO/TD No 942, 7.31-32
and von Storch, H., H. Langenberg and F. Feser, 1999: Long-wave forcing for regional atmospheric modelling. GKSS 99/E/46 report
That the scheme actually is superior than in the conventional formulation in generating mesoscale variability in a correctly described large scale dynamical environment has been demonstrated by a series of authors, in particular Frauke Feser, 2006: Enhanced detectability of added value in limited area model results separated into different spatial scales. Mon. Wea. Rev. 134(8), 2180-2190. Aso an improved description of the formation of North Atlantic Polar Lows (Zahn, M., H. von Storch, and S. Bakan, 2008: Climate mode simulation of North Atlantic Polar Lows in a limited area model, Tellus A, DOI: 10.1111/j.1600-0870.2008.00330.x) and E Asian typhoons (Feser, F., and H. von Storch, 2008: A dynamical downscaling case study for typhoons in SE Asia using a regional climate model. Mon. Wea. Rev. 136, 1806-1815) as well as other mesoscale storms has been shown - as compared to only laterally constrained simulations.
In 2011, the different lines of analysis of the added value generated in this way was described in a review paper Regional climate models add value. in Bull. Amer. Meteor. Soc. (92: 1181–1192) coauthored by F. Feser, B. Rockel, H. von Storch, J. Winterfeldt, and M. Zahn. After having been implemented in limited area software like WRF, the spectral nudging method finds more an more acceptance and is often utilized.
The same terminology and a similar approach had been earlier published by Waldron, K. M., J. Peagle and J.D. Horel, 1996: Sensitivity of a spectrally filtered and nudged limited area model to outer model options. Mon. Wea. Rev. 124, 529-547. At the time of submitting our manuscript, this paper was unknown to Hans von Storch, as it dealt with issues of weather forecasting, and he learned from it during the review process. Earlier a number of authors (Kida, Sasaki, McGregor) had experimented with nudging spatial means.
It was accepted in the early 1970s that ensembles of global model (GCM) simulations, which differ by miniscule details, such as modified initial values or parameters, would show strong variability reflecting internal chaotic dynamics. This observation was important for both validating models, as simple comparisons of observations and a model simulation would not be conclusive becaused the differences may very well be due to ensemble variations, and for doing numerical experiments on the effect of changing parametrizations or boundary conditions (such as sea surface temperature). As a consequence, the concept of using the concept of testing null hypotheses was introduced and became quickly a stand routine in validating of and experimenting with GCMs.
Not so with regional climate models (RCMs). While the situation is in principle the same - the lateral boundary conditions, which are usually stipulated via "sponge zone, are insufficient to enforce a unique solution in the intgerior, and the chaotic dynamics generate - at least sometimes - different circulations in the interior, which are consistent with the lateral boundary conditions. Thus, validation of and experimenting with RCMs needs the same statsitical analysis as for global models. Ioneers in demonstrating the emrgence of ensm,eble variations in RCMs simulations were Ji and Vernekar in 1997 (J. Clim 10), and Annette Rinke and Klaus Dethloff in 2000 (Clim. Res. 14). But even nowadays, in 2012, the statistical challenge of assessing RCM output has not been acknowledged by many in the RCM community.
However, there is a characteristic difference in the ensemble structure between different trajectories generated by global and regional models. Once two global models trajectories have moved from each other, they will remain at a statistically stationary distance. In case of RCMs, this is different, because the efficiency of lateral boundary conditions vary in enforcing a circulation in the interior. Sometimes they do it really well, and any two trajectories move closer, while in other situations the differences grow, sometimes strongly within a short term and only for a limited time.
The principle was described in this article: Weisse, R., H. Heyen and H. von Storch, 2000: Sensitivity of a regional atmospheric model to a sea state dependent roughness and the need of ensemble calculations. Mon. Wea. Rev. 128: 3631-3642, which unfortunately received little attention so far.
At a later time, I coined the term "intermitted divergence in phases space" for this phenomenon. (von Storch., H., 2005: Models of global and regional climate. M.G. Anderson (Ed): Encyclopedia of Hydrological Sciences, Part 3. Meteorology and Climatology, Chapter 32, ISBN: 0 471-49103-9, p 478-490 DOI: 10.1002/0470848944.hsa035) This intermitted divergence in phase space is strongly suppressed, when spectral nudging is employed (see Weisse and Feser, 2003, Coastal Eng. 48.)
Originally introduced by Klaus Hasselmann in a rather general concept, later expanded to Principal Interaction Patterns (PIPs, see Hasselmann, 1988, I derived a simplified but workable version, which nowadays goes with the name of Principal Oscillation Patterns (POPs; see von Storch et al., 1988. The basic idea is to identify a low-dimensional linear subspace, which is dynamically closed (interacting with other dynamcis only through links, which are suitably described by statistical terms). The PIPs represent a very general, and elegant, approach, while the POPs reduced this to the case of 2-dimensional subspace, within which oscillatory modes descrrbe most of the variation. Both, PIPs and POPs incorporate a predictive potential.
PIPs have never been implemented in its full generality; POPs have been successfully applied to the Madden & Julian Oscillation and to ENSO, among other phenomena. A review dates from 1995. The original FORTRAN code from about 1991 has been preserved (thanks to Silvio Gualdi),to gether with the manual (by Fergus Gallagher, Reiner Schnur, Gerhard Hannoschöck and myself) from those days.Hasselmann, K., 1988: PIPs and POPs: The reduction of complex dynamical systems using Principal Interaction and Oscillation Patterns. J. Geophys. Res. 93, 11015-11021
The "hockeystick"-episode was a tough time, also interesting in terms of the social process called science. We have summarized how we perceived this time in a comment on the weblog of nature. The problem was that the scienbtist behind the hockeystick used the unfortunate declaration that the hockeystick would be "it", by the IPCC (TAR) as a tool to block others who would come forward with different results; in particular borehole temperature people suffered.
For the debate about detecting and attributing man-made climate change the hockeystick was unnecessary, as was demonstrated by our later analysis (Rybksi et al.) in 2006, which employed various reconstructions; also formal detection and attribution studies had solved the problem long before (e.g., IDAG, 2005: Detecting and attributing external influences on the climate system. A review of recent advances. J. Climate 18, 1291-1314; or, much earlier Hegerl, G., H. von Storch, K. Hasselmann, B.D. Santer, U. Cubasch, P.D. Jones, 1996: Detecting anthropogenic climate change with an optimal fingerprint method. - J. Climate 9, 2281-2306)
The "skeptics" were keen on deconstructing the hockeystick because they considered it as a key argument "pro" man-made causes. However, it was not, and the fact that the hockeystick methodology was flawed did not imply that the question about man-made causes would be entirely open again. (Of course there is some doubt left, because of significant uncertainty in the level of natural variability - but this doubt is small.) This demonstrates very clearly that overselling does not pay; instead it endangers the credibility of the whole community.
The significance of hockeystick business is illustated by the fact that it brought me an invitation to show up as a witness first at the National Research Council and later at the Subcommittee on Oversight and Investigations of the Committee on Energy and Commerce, U.S. House of Representatives. (Hearing "Questions Surrounding the ‘Hockey Stick’ Temperature Studies: Implications for Climate Change Assessments" on 19. July 2006)
The key-paper of this effort was
von Storch, H., E. Zorita, J. Jones, Y. Dimitriev, F. González-Rouco, and S. Tett, 2004: Reconstructing past climate from noisy data, Science 306, 679-682, 22 October 2004 (Sciencexpress, doi 10.1126/science.1096109)
which was commented to by Osborn and Briffa, and by S. Rahmstorf.
A kind of final word from our side is provided by
von Storch, H., E. Zorita and J.F. González-Rouco, 2009: Assessment of three temperature reconstruction methods in the virtual reality of a climate simulation. International Journal of Earth Sciences (Geol. Rundsch.) 98:67–82 DOI 10.1007/s00531-008-0349-5 .
The ongoing trans-disciplinary cooperation between the theoretical sociologist Nico Stehr and the climate scientist Hans von Storch has also resulted in a joint book, a first issue published in German in 1999 (and later in Chinese), and a second in 2010 in English, German, Slowenian, Russian, and Tchechian:
So far such reports on the scientifically legitimate knowledge about climate, climate change and climate impact have been published:
A short version of BACC-I is given by
Reckermann, M., H.-J. Isemer and H. von Storch, 2008: Climate Change Assessment for the Baltic Sea Basin. EOS Trans. Amer. Geophys. U., 161-162
An analysis of the recognition of the BACC report among regional stakeholders is provided by
Bray, D., 2010: Baltic Climate Scientists’ Assessment of Climate Change and Climate Science in the Baltic Sea Basin, International BALTEX Secretariat Publication 48, ISSN 1681-6471, 79 pp.