Downloaded 09/24/21 10:19 PM UTC on Scientific Accountability and the Questioning and Discussion of Scientific Papers and Professionalism Operations

letters to the editor

Sverre Petterssen s Autobiography On Scientific Accountability and Professionalism US. Trends in Streamflow and Precipitation: Using Societal Impact Data to Address an Apparent Paradox

Sverre Petterssen's Autobiography winds over Europe—what we now call the "jet stream"—seems to have been held in tight security by Sverre Petterssen, who died in England at the end upper-air forecasters in England in the early 1940s, of 1974, was an extremely active meteorologist even among other Allied weather forecasters.) throughout his life, in his native Norway, in England Petterssen recounts his feeling that this warning, during World War II, and both before and after the war although based on his English experiences in 1942 and in the United States. He wrote his autobiography in 1943, was not taken seriously by his audience. These English, but arranged for its publication in Norwegian winds were soon experienced by the new B-29 aircraft, as Kulingfra Nord, en vaervarslers erindringer, trans- as has been told by Plumley (1994) and Bryson (1994), lated as Gale from the North, a Forecaster's Remem- who to their great credit first forecast them in that area, brances (Petterssenn 1974). Published in 1974 by on the basis of little data. H. Aschehoug and Company of Oslo, it has not been Quite apart from such human details of meteoro- generally available in this country. Professor Arnt logical interest, Petterssen seems to have written with Eliassen of Oslo was kind enough to locate a copy for complete candor on all subjects, making this a very me, which has been deposited in his name at the compelling biography, and I wish the Historical Com- Massachusetts Institute of Technology Library.1 mission success in their effort to make it available to This site is convenient for interlibrary loan. Petterssen us all. served as head of the meteorology department at the Massachusetts Institute of Technology for two years following the departure of C.-G. Rossby. References During World War II Petterssen was active in many ways, many of which have been described by Bryson, R., 1994: The discovery of the jet stream. Wisconsin Acad- Johannesen (1975) and Bundgaard (1979). An event emy Rev., xx, 15-17. of historical interest described in Kulingfra Nord, but Bundgaard, R. C., 1979: Sverre Petterssen, weather forecaster. not treated by Johannesen and Bundgaard, was a short Sverre Petterssen memorial lecture. Bull. Amer. Meteor. Soc., series of lectures that Petterssen gave in Hawaii in the 60, 182-195. Johannessen, K. R., 1975: Sverre Petterssen, 1898-1974. Bull. fall of 1944, after his duties for the invasion of Europe Amer. Meteor. Soc., 56, 892-894. had eased off. He recalls that in those talks, designed Petterssen, S., 1974: Ruling fra Nord, en vaervarslingers to prepare a small audience of forecasters for the com- erindringer. H. Aschehoug, Oslo, 312 pp. ing attack on the Japanese home islands, he mentioned Plumley, W., 1994: Winds over Japan. Bull. Amer. Meteor. Soc., that strong westerly winds of as much as 400 km h_1 75, 63-68. should be expected in winter. (Knowledge of strong

NORMAN PHILLIPS '77 Massachusetts Avenue, Cambridge, MA 02139-4307. MERRIMACK, NEW HAMPSHIRE

Bulletin of the American Meteorological Society 1433

Unauthenticated | Downloaded 09/24/21 10:19 PM UTC On Scientific Accountability and the questioning and discussion of scientific papers and Professionalism operations. No matter how poor we think the science is, we should strive to understand what is being done Scientific accountability and professionalism is and give advice as to how to improve the work. The sometimes lacking in our participation in scientific proper time for sightseeing is before or after the con- conferences, particularly the international ones. ference and with the appropriate leave time taken. Among the worst offenders are some senior scientists Some U.S. scientists are members of international and professors from our research institutions and panels of experts or other scientific panels. They in universities. particular have a responsibility to learn as much about I say this as I complete two terms as chairman of the field as possible because they are the U.S. repre- the World Meteorological Organization Executive sentative on the panels. Part of their job is to review Panel of Experts on the Physics and Chemistry of the status of the field and interact with their foreign Clouds and Weather Modification Research. It was my colleagues. responsibility to organize, as program chairman, the All of this has made me think about scientific ac- Sixth and Seventh Conferences on Weather Modifi- countability and professionalism. Perhaps we should cation, one held in Paestum, Italy, and the most recent teach such a course in our universities, as they do in in Chiang Mai, Thailand. In addition, I had attended our engineering school's curriculum. Certainly, our the Fifth Conference in Beijing, where a Chinese sci- research organizations and universities should take entist had expressed to me his dissatisfaction with the action to ensure that their senior scientists and profes- amount of interaction with foreign scientists because sors set a good example for the community. Poor par- of their absence from many of the sessions. What dis- ticipation in scientific conferences is really a rip-off turbs me is the practice of some scientists from the of public funds. This lack of effort is certainly slow- United States or other countries to attend these con- ing up progress in our science. We have little time to ferences, perhaps present a paper or chair a session, waste in some very important areas, water resources and then disappear from the meeting. The mind-set of and cloud seeding being examples. the scientist may be that the presentation is the Although my comments have been centered about scientist's contribution to the conference. In truth, an international scientific conference, I am sure they though, the discussions and constructive criticisms apply to many of our domestic conferences also. And during the scientific sessions are an important com- I am not without blame. I remember playing a golf ponent of the conference, particularly for those scien- game with a senior professor during a conference in tists attending from a country with a poorly developed Seattle. scientific infrastructure. If you have similar thoughts and know of ways to In the most recent conference, the poorest example encourage more responsible scientific participation in of scientific responsibility was the organization by a conferences, please let me ([email protected]) senior scientist from another country of an all-day or the STAC Commissioner of the American Meteo- outing to the northeast of Thailand during the day de- rological Society (Dr. Richard D. Rosen; rdrosen@ voted to hail studies. Other scientists, some from the aer.com) know. United States, joined him on the trip. In my view, scientists being supported by public funds to attend scientific conferences have a responsibility to attend and to contribute to the entire meeting, particularly ones that are international in flavor HAROLD D. ORVILLE and depend on the input of U.S. scientists in the field. DISTINGUISHED PROFESSOR EMERITUS The field of weather modification needs constructive SOUTH DAKOTA SCHOOL OF MINES AND TECHNOLOGY criticism and the active participation of scientists in RAPID CITY, SOUTH DAKOTA

1434 Vol. 80, No. 7, July J 999

Unauthenticated | Downloaded 09/24/21 10:19 PM UTC U.S. Trends in Streamflow and A recent study (Pielke and Downton 1999, manu- Precipitation: Using Societal Impact script submitted to J. Climate) offers an analysis that Data to Address an Apparent Paradox helps to address the apparent paradox. Pielke and Downton relate trends in various measures of precipi- Recently, Lins and Slack (1999) published a pa- tation with trends in flood damage in the United States. per showing that in the United States in the twentieth The study finds that the increase in precipitation (how- century, there have not been significant trends up or ever measured) is insufficient to explain increasing down in the highest levels of streamflow. This follows flood damages or variability in flood damages. The a series of papers showing that over the same period study strongly suggests that societal factors—growth "extreme" precipitation in the United States has in- in population and wealth—are partly responsible for creased (e.g., Karl and Knight 1998a; Karl et al. 1995). the observed trend in flood damages. The analysis The differences in the two sets of findings have led shows that a relatively small fraction of the increase some to suggest the existence of an apparent paradox: in damages can be associated with the small increas- How can it be that on a national scale extreme rainfall ing trends in precipitation. Indeed, after adjusting dam- is increasing while peak streamflow is not? Resolving ages for the change in national wealth, there is no the paradox is important for policy debate because the significant trend in damages. This would tend to sup- impacts of an enhanced hydrological cycle are an area port the assertion by Lins and Slack (1999) that in- of speculation under the Intergovernmental Panel on creasing precipitation is not inconsistent with an Climate Change (Houghton et al. 1996). absence of upward trends in extreme streamflow. In There does exist some question as to whether com- other words, there is no paradox. As they write, paring the two sets of findings is appropriate. Karl and Knight (1998b) note that We suspect that our streamflow findings are consistent with the precipitation findings of As yet, there does not appear to be a good Karl and his collaborators (1995, 1998). The physical explanation as to how peak flows reported increases in precipitation are modest, could show no change (other than a sampling although concentrated in the higher quantiles. bias), given that there has been an across the Moreover, the trends described for the extreme board increase in extreme precipitation for 1- precipitation category (> 50.4 mm per day) are to 7-day extreme and heavy precipitation not necessarily sufficient to generate an in- events, mean streamflows, and total and annual crease in flooding. It would be useful to know precipitation. if there are trends in 24-hour precipitation in the > 100 mm and larger categories. The term Karl's reference to a sampling bias arises because "extreme," in the context of these thresholds, of the differences in the areal coverage of the Lins and may have more meaning with respect to Slack study and those led by Karl. Lins and Slack fo- changes in flood hydrology. cus on streamflow in basins that are "climate sensi- tive" (Slack and Landwehr 1992). Karl suggests that Karl et al. document that the increase in precipita- these basins are not uniformly distributed over the tion occurs mostly in spring, summer, and fall, but not United States, leading to questions of the validity of in winter. H. Lins (1999, personal communication) the Lins and Slack findings on a national scale (T. Karl notes that peak streamflow is closely connected to 1999, personal communication). While further re- winter precipitation and that "precipitation increases search is clearly needed to understand the connections in summer and autumn provide runoff to rivers and of precipitation and streamflow, in this letter we re- streams at the very time of year when they are most port the results of a recent study on the relationship of able to carry the water within their banks. Thus, precipitation and flood damages. This letter seeks to we see increases in the lower half of the streamflow address the apparent paradox from the perspective of distribution." societal impacts. We suggest that an analysis of the Furthermore, McCabe and Wolock (1997) suggest relationship of precipitation and flood damages pro- that detection of trends in runoff, a determining fac- vides information that is useful in developing relevant tor in streamflow, are more difficult to observe than hypotheses and placing the precipitaton-streamflow trends in precipitation: "the probability of detecting debate into a broader policy context (cf. Changnon trends in measured runoff [i.e., streamflow] may be 1998). very low, even if there are real underlying trends in

Bulletin of the American Meteorological Society 1435

Unauthenticated | Downloaded 09/24/21 10:19 PM UTC the data such as trends caused by climate change." climatological aspects of flooding, but also in a better McCabe and Wolock focus on detection of trends in understanding of the societal aspects of flood damage. mean runoff/streamflow, so there is some question as (See Pielke 1999 for further discussion.) to its applicability to peak flows. If the findings do hold at the higher levels of runoff/streamflow, then this Acknowledgments. We thank T. Karl, H. Lins, and an anony- would provide another reason why the work of Lins mous reviewer for useful comments on an early draft of this letter. The National Center for Atmospheric Research is sponsored and Slack is not inconsistent with that of Karl et al., by the National Science Foundation. as it would be physically possible that the two sets of analyses are complementary.1 In any case, an analysis of the damage record shows References that at a national level any trends in extreme hydrological floods are not large in comparison to the Changnon, S. A., 1998: Comments on "Secular trends of precipi- growth in societal vulnerability. Even so, there is a tation amount, frequency, and intensity in the United States." documented relationship between precipitation and Bull. Amer. Meteor. Soc., 79, 2550-2552. flood damages, independent of growth in national Houghton, J. T., L. G. Meira Filho, B. A. Callander, N. Harris, A. Kattenberg, and K. Maskell, Eds., 1996: Climate Change population: as precipitation increases, so does flood 1995: The Science of Climate Change. Cambridge Univer- damage. From these results it is possible to argue that sity Press, 572 pp. interpretations in policy debate of the various recent Karl, T. R., and R. W. Knight, 1998a: Secular trend of precipita- studies of precipitation and streamflow have been mis- tion amount, frequency, and intensity in the United States. Bull. leading. On the one hand, increasing "extreme" pre- Amer. Meteor. Soc., 79, 231-242. cipitation has not been the most important factor in , and , 1998b: Reply. Bull. Amer. Meteor. Soc., 79, 2552-2554. documented increase in flood damage. On the other , , and N. Plummer, 1995: Trends in high-frequency cli- hand, evidence of a lack of trends in peak flows does mate variability in the twentieth century. Nature, 377,217-220. not mean that policy makers need not worry about Lins, H., and J. R. Slack, 1999: Streamflow trends in the United increasing precipitation or future floods. Advocates States. Geophys. Res. Lett., 26, 227-230. pushing either line of argument in the policy arena risk McCabe, G. J., Jr., and D. M. Wolock, 1997: Climate change and misusing what the scientific record actually shows. the detection of trends in runoff. Climate Res., 8, 129-134. Pielke, R. A., Jr., 1999: Nine fallacies of floods. Climate Change, What has thus far been largely missed in the debate is 42 (2), 413^38. that the solutions to the nation's flood problems lie not Slack, J. R., and J. M. Landwehr, 1992: Hydro-Climatic Data only in a better understanding of the hydrological and Network: A U.S. Geological Survey streamflow data set for the United States for the study of climate variations, 1874- 1988. U.S. Geological Survey Rep. 92-129, 193 pp. [Avail- able from USGS, Reston, VA 20192.] 'An anonymous reviewer offers a similar explanation: "It might be that there are trends in annual peak flow, but that the trends are masked by natural variability. The only difference between the Karl results and the Lins and Slack results may be that natural variability is larger for annual peak streamflow than it is for the precipitation data examined by Karl and collaborators. Thus, ROGER A. PIELKE JR. AND MARY W. DOWNTON there may not be an inconsistency in what the data are indicat- ENVIRONMENTAL AND SOCIETAL IMPACTS GROUP ing. The two sets of data simply have not been examined in a NATIONAL CENTER FOR ATMOSPHERIC RADIATION similar manner." BOULDER, COLORADO

iy tx

1436 Vol. 80, No. 7, July J 999

Unauthenticated | Downloaded 09/24/21 10:19 PM UTC