For more information, purchase options, and for other versions (if available) please visit
Contents | Data for Science and Society: The Second National Conference on Scientific and Technical Data | U.S. National Committee for CODATA | National Research Council Chapter 6: The Intergovernmental Panel on Climate Change and the Futures Market: Common Climate Data Management Requirements? | Data for Science and Society: The Second National Conference on Scientific and Technical Data | U.S. National Committee for CODATA | National Research Council

U.S. National Committee for CODATA
National Research Council
Interdisciplinary and Intersectoral Data Applications: A Focus on Environmental Observations


The Intergovernmental Panel on Climate Change and the Futures Market:
Common Climate Data Management Requirements?

Thomas Karl

     The Intergovernmental Panel on Climate Change (IPCC) was established in the late 1980s to look at the issue of how man may be affecting the global climate. As part of that initiative, one of the important activities was to try to establish whether the climate was indeed changing. As you can imagine, this exercise required going back to climate records and trying to establish their credibility, to determine whether or not what we think we see in the climate record was really reflective of what the climate was doing.

     Figure 6.1 shows the number of billion-dollar U.S. weather disasters from 1980 through 1999. Each of the icons reflects a particular type of weather disaster, such as hurricanes, tornadoes, major blizzards, heavy rainstorms, forest fires, and flooding. Quite a number of these events have happened over the last couple of decades, and the Southeast seems to be particularly vulnerable. Obviously, this has caught the attention of a number of people, including those who actually stand to lose a considerable amount of money through these particular events.

Figure 6.1


     What did the IPCC and the weather risk financial markets have in common? In order to look at these extreme weather events and weather events in general, the question of whether they are changing in terms of their frequency was raised. This is an important question if we are worried about risk. If we actually develop some kind of financial tool to hedge against this risk, the weather risk industry wants to make sure it can settle against a claim using data in which it has confidence.

     Both the IPCC and the weather risk financial markets share the need for high-quality homogeneous climate records. What do we mean by homogeneous? These are climate records that truly reflect changes and variations in climate. If we are interested in looking at whether the climate has really changed in a specific location, we may look at climate records and find that the change is due to the growth of a local urban environment around the weather station and not to the broader climate itself. Or it may reflect an instrument change and not a real shift in climate. For example, if you look at U.S. national temperatures for 1900-1999, you can see many changes and variations. It is critical to know which of these variations do or do not reflect regional and large-scale changes.

     Some of the challenges that the IPCC faces include inaccessible data. You heard Dr. Baker this morning talk about trying to make data more accessible, which essentially means putting them into some kind of computer format. Inadequate information about measurement methods is another major factor because once you get these data in digital formats, the issue is what they really represent. Another challenge is the changing environment around observing sites. We are talking about long-term data, with regard to land observations. Certainly there have been a lot of land-use changes over the years. An important characteristic that is stated in IPCC 1990, was emphasized in the 2000 report, and is being worked on now focuses on establishing internal consistency among variables to ensure that the large-scale picture of changes makes physical sense.

Figure 6.2


     Figure 6.2 illustrates precipitation measurements for a number of different countries and the inhomogeneities in precipitation measurements. For example, in Finland a new gauge was introduced in 1909, and again in 1981, for liquid precipitation. The estimated change in precipitation due to the new instrument was a 4 percent increase; for solid precipitation, a 20 percent change, resulted in 1909, and another 5 to 13 percent in 1981. So to use these records, you have to be extremely careful about the history in which the observations were made. Once you understand this history, you have to try to evaluate the effect of these instrument changes on the measured precipitation amount.

     In 1995, the IPCC also focused on the global and regional environment, looking at other variables besides temperature and precipitation, such as snow cover. The IPCC captured the public's attention by focusing on the question of detection and attribution of climate change. It is one thing to detect climate change by saying that the climate is changing because it is statistically unusual. However, attributing a particular change to some particular causes requires details of the patterns of change. So the IPCC was very much involved in looking at not just global-scales but regional changes and getting the best information it could on regional time scales. A special interest also was taken in extreme weather and climate events. It was no longer a matter of just global temperature changes. Things like changes in extreme events, such as how frequently were we seeing heavy rainfall, record-breaking temperatures, or hurricanes, were of utmost concern.

     This presented new challenges. It was very clear to the IPCC that it was a lot easier to try to establish the homogeneity of means and averages over long periods of time than look at extreme temperatures. In recognizing many of the problems involved in trying to make these measurements, a set of principles for climate monitoring was developed. In fact, a resolution was passed at the Fifth Conference of the Parties meeting convened every 1 to 2 years in which all nations get together and assess greenhouse gas emissions to fix this problem. Most of you probably remember the Kyoto meeting, which established a set of emission reduction targets for greenhouse gas emissions. The principles and guidelines addressed things such as changes in data collection methods. Consideration was given to metadata, related to the full documentation of observing system characteristics including operating procedure algorithms used to process the data.

     For IPCC 2000, there is an even greater emphasis on extreme events and, as I mentioned earlier, a renewed emphasis on consistency among variables. There is new recognition that we need to focus on extreme events to understand how our climate is changing.

     It is my understanding that about 1 to 2 billion dollars in "weather risk" trades have been made just over the past year. Why? The notion of trying to hedge against possible losses due to the vagaries of climate and weather is important for consistent profit margins. A variety of contracts are now tied to heating and cooling degree-days. There is also interest in other variables such as precipitation. The issues being faced today are very similar to those that the IPCC has been looking at over the last decade or more. What is the present status of existing observations? Are there random errors in these measurements? Are there systematic errors? The IPCC is always faced with trying to get the latest data every 5 years for an assessment to bring the figures up-to-date, but the weather risk community wants updates for 60-day settlement time periods. This community is looking for some official and reliable documentation because it has to settle in a short period of time. So it is pushing the time horizon at which we can look and trying to address some of these systematic errors. This period is much shorter than we ever thought.

     I think that some very similar interests have developed with respect to stewardship of the data. Clearly, this is an issue that has captured the attention of the weather risk community. Real-time access is something that this community is pushing scientists to address, and it challenges our ability to rapidly identify and correct systematic errors. This is a whole new sector that needs some education about how to use these data, and of course, the education will go both ways. Whereas before this was a scientific question in understanding the global environment, and it was difficult to identify just how much money was involved with respect to the effects of climate change, some very different motivations with very large resource considerations are now at stake.

Copyright 2001 the National Academy of Sciences