John E. Lewis and Eric C. Wood
[The following presentation was initially offered as a keynote address at the Second International Congress of Arctic Social Sciences, (ICASS II) Rovaniemi, Finland, 1995. It was published in Unity and Diversity in Arctic Societies, in 1996 by the International Arctic Social Sciences Association. Reprinted with permission of the authors and publisher.]
Over forty years ago, C.P. Snow expressed concern over the separation of the two scientific cultures - the natural and the social - and urged for more cooperation in their research. Now, more than ever, there is an urgency to bring the two together. However, problems exist when trying to foster cooperation mainly on questions of the underlying scientific paradigms and the testing of proofs. These factors have traditionally separated the two cultures. The root of the conflicts arises because, as Blaikie (1985) puts it, "both sets of disciplines tend to have different conceptions about the domain and status of proof in the pursuit of knowledge....". The main problem seems to derive from the positivist expectation that (e.g.) the explanation of social phenomena should be tested in the same way as most natural science ones."
Both Kuhn ( 1970) and Phillips ( 1971 ) have raise pointed concerns . If we are going to benefit from the "cooperation of natural and social scientists", then we are going to have to get beyond the constraints of the paradigms of the individual disciplines. Researchers will have to be willing to accept the validity of multiple paradigms, which appears to go against the norms of science as we have defined it. Or perhaps, the only way this integration of the disciplines will work is for all partners to work under a totally new paradigm, which might be called a "universal paradigm shift". These are not new thoughts but given the preceding remarks as a backdrop, let us now turn our attention to a discussion of global change.
Institutional impediments, lack of political will and socio-economic forces have placed major strains on the sub-Arctic and Arctic environments. Global change is occurring at an ever increasing rate, and this change is nowhere felt more than in the delicate environmental systems within both the oceanic and terrestrial regimes for the northern latitudes. These regions face a myriad of environmental problems that at times appear to be overwhelming - depleting of fishing stocks and decline in marine mammal populations; decreased habitat for caribou/reindeer herds; continual pressures of mineral exploration and its social consequences on native groups; and the socio-economic impacts of climatic change. And these serve as only a few examples.
The bottom line in solving these problems relies on political will and will ultimately require political decisions to attain environmental solutions. However, for these problems to reach the political forum with the chance of achieving a successful outcome, we feel it will require a much greater cooperation between the two scientific cultures. Only a comprehensive presentation along with an increased lucidity and veracity of the causes and implications of these environmental problems will serve to convince the public and the political decision makers of the serious concerns that face society. This means a newfound understanding must be obtained of the concepts and methodologies that each of our scientific disciplines have to offer.
As a climatologist, I shall address one area with which I am familiar; i. e., the problem of climate change as an example of global change and how these notions fit it into the realm of the human dimension. An aspect of 'global warming' will be used as the example to illustrate some of concepts which have the potential for fostering intellectual and informational integration among physical and social scientists.
The specific topics that we shall discuss:
1) a general definition of global change;
2) recent findings about climatic change in the northern latitudes;
3) dealing with the human dimensions; and 4) an example of some emerging tools for analyzing global change.
There is a general lack of consensus in the scientific community concerning both a comprehensive definition and the conceptual scope of global change. Global environmental change is not a new idea. This concept has been a topic of academic and public interest for many years. George Perkins Marsh writing in the mid-nineteenth century published the now classic work entitled Man and Nature or The Earth as Modified by Human Action. Marsh in his works emphasized the human impacts on the natural landscape (Marsh 1965).In the year 1864, he challenged the notion of an inexhaustible earth. More recently with the publication of Man's Role in Changing the Face of the Earth ( 1956) and with the environmental movement of the 60's and early 70's, there was a returned emphasis to the concept of global environmental change.
In the recent work entitled Global Environmental Change: Understanding the Human Dimensions, published in 1992 by the U.S. National Research Council, global change is defined as 'alterations in the natural (e.g., physical or biological) systems whose impacts are not and cannot be localized' (Sternetal. 1992: p.25). Global change research initially focused almost solely on the physical and biological sciences, but in the last eight to ten years, there has been more recognition of the human dimensions and the essential role that social science must play in resolving global environmental problems. As evidence, the three problems that have been given greatest attention climate change, ozone depletion and loss of biodiversity - are all anthropogenic in origin (Stem et al. 1992). There is an increasing realization that we need to have a much better understanding of how human activities combine with natural events to produce global changes. We must explore causes and consequences of how human, interacting within the social systems, affect and are affected by global change (CIESIN 1992). The processes of global change tend to be highly non-linear and are characterized by human responses that can have positive or negative feedbacks. If solutions, or even just an increased comprehension to many of these environmental problems, are to be obtained, a more integrated natural/social science approach must be adopted.
Global environmental change can have two types of meaning: 1) systemic and 2) cumulative (Fig. 1). Global systemic changes need not be caused by global scale activity; only the physical impact is experienced at the global scale (Turner, et al. 1990). The pollutants of suIfur dioxide and the oxides of nitrogen are emitted on a local scale; but through the mechanism of long range atmospheric transport their effects are evident on the global scale in the form of acid rain. Whereas, the cumulative type of global change refers to the areal or substantive accumulation of localized change which becomes global if it happens on a worldwide scale. In other words, small effects accumulate and exceed some spatial or temporal threshold thus producing a global problem. This is what has been called " the tragedy of small increments". Examples of problems of this latter type are the loss of biodiversity and the gradual conversion of forest, grassland and wetland.
Changes as mentioned in the previous paragraph, are produced by human actions and these actions are considered as "driving forces". Four general categories of activities have been identified as human driving forces: 1) population change, 2) wealth/poverty levels, 3) economic structures and 4) technology (Meredith et al. 1994). How these human driving forces are incorporated into a conceptual framework of global change will be discussed in a later section.
It is difficult to summarize the vast literature even for the last 10 years that has been written on climatic change (Houghton et al. 1992). In this section we shall highlight a few of the more recent findings which have implications for northern environments.
Many arguments and questions surround the area of climatic change, especially the topic of global warming. Global temperature data sets generally confirm that surface warming between 0.3 and 0.6°C has occurred in the last 100 years (Karl 1993), and the ten warmest years of the century have all been since 1979, and include 1993. However, regional studies fail to confirm any worldwide trends and these studies tend to show a complex spatial and temporal pattern of temperature trends. In the higher northern latitudes, atmospheric temperature trends are found to be seasonally and spatially variable (Kahl e tal. 1993) especially for Arctic regions; and these temperature changes are less significant than for tropical regions. Temperatures in the North Atlantic Ocean (within the 300m to 600m depth) actually show a cooling trend (Antov 1993). Other results indicate a strong positive correlation with NW Atlantic sea surface temperatures and Arctic sea ice conditions.
If we use other climatic parameters to assess the climatic change in the north the situation is even more hazy. Precipitation data sets tend to be more sparse and noisy than the temperature records for Arctic and sub-Arctic regions and may be seriously affected by measurement errors. Therefore, it is difficult to arrive at any firm conclusions about precipitation trends. For sea ice, there is some evidence of decreased amounts in the Arctic regions over the last four decades with the decrease especially significant for the summer seasons (Chapman and Walsh 1993).
When we turn our attention, briefly, to studies which focus on potential impacts as a result of proposed temperature change, two recent studies concerning changes in northern ecosystems have interest:
(1) An analysis using a modified Budyko model (a very simplified climate/biosphere model) predicted that all boreal forest vegetation classes will shrink and will be replaced by taiga and temperate forest classes moving northward (Morserud el al. 1993).
(2) As a contrast to the previous modeling study, results using paleoecological evidence suggest how changes in vegetation may occur in the future for the boreal forest/tundra area.
Results indicate that for a warm period (4000 to 5000 years BP), the vegetation changed from tundra to close canopy black spruce in a matter of 150 years (MacDonald et al. 1993). The tacit inference here is that the past historical warm period had the same overall climate conditions as those conditions producing the temperature changes presently predicted for the northern regions. The effect these vegetation changes would have on the animal population and the people living in these environments is unknown.
These two studies contrast how information is obtained on what may happen in the future - the latter by developing future climatic scenarios based on historical situations and the former with the use of physical process model to simulate future climate (Lewis 1989). By far, the process mode approach has been the bulwark for assessing climate change and making predictions on future climates. The success of these models in simulating the global climate system hinges on the models' ability to represent the physical and chemical processes and how the important feedbacks with in the system are described and parameterized (AES 1994). These general circulation models (GCM) are areas of very active research within the climate community and are continually improving in their description of the climate system on a global scale; however, substantial concerns still exist about the reliability, representativeness and spatial resolution of these models used in the analysis of climate change.
The U.S. Global Change Research Program recently published a report entitled Forum on Climatic Modeling ( 1995) which provides the latest consensus conclusions on model prediction for future climate conditions. The models' conclusions are placed into three categories: Very Probable, Probable and Uncertain. In this paper we list only the ones that have direct implications for the north. This is not to infer that other systemic types of global change, such as volcanic eruptions or surface cover/land use changes, will not be important contributors to global change in the future.
1) Global temperatures will increase between 0.5 - 2.0°C by 2050.
2) Northern Hemisphere sea ice will decrease; however some areas will have expansion .
3) Arctic land areas will show increased warming and reduced snow cover. (The caveat here is how well the models describe the poleward transport of heat.)
4) Global sea level will rise at an increasing rate resulting in a higher sea level of 5 - 40 cm by 2050.
1) High latitude precipitation will increase.
2) TheNorth Atlantic Ocean will experience warming at a slowerrate than the global average.
1) Changes in climatic variability will occur.
2) Regional scale ( 100-2000 km) climate change will be different than the global average.
3) Details of climate change over the next 25 years are uncertain.
4) Biosphere-climate feedbacks are expected but whether these feedbacks will amplify or moderate climate change is uncertain.
With uncertainty the major feature of future climatic conditions, and the climate system is only one component, how do we deal with the myriad of concepts and processes within the total scope of global change? In recognition of this problem, the Consortium for International Earth Science Information Network (ClESIN ) conmissioned a group of scientists to develop a framework illustrating the key social/human dimensions that contribute to global change.
The result of this effort was the Social Process Diagram (fig. 2) which serves as a dynamic tool for both natural and social scientists in fostering an increased understanding of global change. The form of the diagram facilitates discussion and research concerning the human dimensions of global change.
The diagram consists of three elements: the structure, the connections, and the dynamics. 1) Structure - The diagram consists of seven building blocks in which six define the human social systems and one the natural systems.
2) Connections - The connections provide the links amongthe building blocks and help createthe process interrelationships. These links define the fundamental driving forces producing environmental change.
3) Dynamics - The dynamics are implicit within the diagram but essential for the understanding of mechanisms of global change.
Driving forces must be considered from both a spatial and temporal perspective. In other words, human interactions occur within certain geographic locations and over certain geographic periods which both contribute to how the diagram is used (CIESIN 1992:24).
An obvious question is how can we use the social process diagram to formulate a research strategy and also point directions for providing all interdisciplinary approach to problem solving. As was noted in the discussion on climatic change effects, sea level rise seems to be a possible consequence of global warming and of particular interest to groups inhabiting northern coastal regions. The social process diagram provides pathways for investigating the sea level rise problem and can contribute to the initial understanding of the environmental/human inter-relationship. The use of the diagram is not to oversimplify the problem but to act as a point of departure from the natural factors to the human dimensions. In other words, it serves as a signpost for research direction and a vehicle for the development and the testing of hypotheses.
In the context of this paper, we shall only describe a few of the process interractions which emphasize more of the physical-human pathways. This section is essentially paraphrased from the case study 'Global Warming and Sea Level Rise' cited in Pathways of Understanding (CIESIN 1992: 36-39). Using this as an example of environmental global change, the social process diagram (fig. 3) displays the structural components for all eight interactions within the realm of human dimensions.
Path A: Environmental Processes -> Economic Systems.
This pathway incorporates the rate of increase in sea level rise and the decrease in usable land. For developed countries, the change can be anticipated and actions might be taken to modify or change the economic activities impacted by the environmental global change. Sea level rise would induce changes he land prices and activities, and in turn effect the pattern of consumption and land development. A few of the major questions concerning global change raised by this pathway are: what level of risk are we willing to accept before any anticipatory action is initiated? How would the risk of land inundation be internalized in the market economy? Or what are the anticipated rates of growth or decline which are caused by this environmental change?
Path B-C-D. Environmental Processes -> Population & Social Structure -> Factors of Technology -> Economic Systems.
A very complex pathway but, by far, this pathway depicts the most interrelated human processes in association with environmental global change. Changes in sea level which decreasethe amount of land available would produce several apparent effects. People occupying inundated areas would have to relocate inland producing changes in population distribution. These changes would then influence the resource base and land use through path C. Agriculture, fishing and recreation/tourism are some of the activities impacted by these land use changes.
Path H. Environmental Processes -> Preferences & Expectations -> Political Systems.
This connection may be characterized as mainly superficial non-economic factors which influence ecological distributions and brings into play how society perceptually and politically responds to these changes. For instance, as sea level rise occurs more pressure from the public may be felt to preserve coastal wetlands. Since many wetlands are preserved for wildlife protection and public usage, social preference would exert direct pressure on the political system. The land use changes would potentially instigate a political discussion over a whole range of non-economic impacts and how society might respond. In addition, Conservation practices that may be initiated would produce conflicts in other areas.
"Protection of economic assets may run counter to other objectives, and these opposing forces should be quantified by physical impact: what ecological systems are threatened doubly by sea level rise on the one hand and econom i cal Iy motivated protection policies on the other?" (ClESIN 1992: 39).For native communities which may only peripherally participate in the global market economy, what impacts are addressed on the social process diagram? One outcome is coastal flooding which could have profound ecological effects on habitat loss or its redistribution. These changes would impact wildlife areas and changethe land/aquatic animal relationship which in turn, could generate repercussions for native activities and their livelihoods. These human dimension problems could assume the form of population redistribution and/or dysfunction he the social structure due to village relocation .
Decision theory is not a new methodology but the integration of it into a spatial framework is. The merging of GIS with resource allocation decision making in order to provide a decision support tool is an excitingand challenging means for analyzing global change problems. Topics within this broad field range from dealing with the uncertainty and errors in the spatial domain including the use of fuzzy logic and Bayesian probability mapping to multi-objective / multi-criteria decision modeling. (See Carver 1991 as an example of the latter technique). The IDRISl group at Clark University has been one of the leaders in this area both in the realm of education for UNITAR and research development.
For this paper we shall provide a simple but illustrative example of incorporating uncertainty, risk assessment and Bayesian probability into a global chalige problem . We have already discussed the consequences of sea level rise within the context of the Social Process Diagram. In this section of the paper, we shall continue with this example, not as a sign of its high priority or overwhelming global importance as an environmental problem, but for pedagogic reasons.
A digital elevation model (DEM) of an area west of Prudhoe Bay on the North Slope of Alaska was obtained from the USGS. The data set, which encompasses 3 arc seconds (1:250000), is for the Beechey Point quadrangle. The original area was reduced by one-half and resampled to a consistent 30m grid (fig. 4) . Given the existing topography , what effect will sea level rise have on this region? If we chose a prediction estimate of sea level rise from one the GCM's results, we can then map the area inundated by this rise. As mentioned previously, there is much discrepancy in the model predictions so we have chosen a value on the high side of 1 .9m rise by the year 2100 for representative purposes.
In a standard analysis, the assumption is that the data is error free. Using this assumption then any land less than 1 .9m will be flooded. (We are not including the effects of or changes in sea ice within the problem matrix.) However, errors exist in the DEM database and there are, as already noted, uncertainty in the predicted value of sea level rise. A RMS value can be calculated for the elevation data but the uncertainty in the GCM model value is more problematic. As best determined from the literature, the uncertainty for the sea level rise of 1.9m is +/- Im. Instead of a hard decision rule with a specified boundary (threshold), there now exists a soft one. Also, a probability for each data point can be calculated which portrays a degree of uncertainty in the decision rule.
Both types of data uncertainty then can be utilized to evaluate the probability that the elevations are exceeded by the sea level rise of I .9m. The resulting map is one of probabilities ranging from 0 to 1 with (e.g.) a data value of .40 indicating that a location has 40% chalice (probability) of remaining above water or conversely, a 60% chance of being flooded. This type of map (fig. 5) presents a much different picture than a map based on a hard decision rule which includes no errors.
It seems somewhat unreasonable that the ocean grid cells are included in the analysis he the same manner as the land areas. Clearly, advance knowledge indicates that the chance of the present ocean area being submerged as a result of higher sea level has to be 100%. Having this prior knowledge would seem to provide some additional advantage in the testing of our hypotheses using this probability approach. The previous probability map serves as initial evidence and then along with the prior probability that the ocean area will have a 100% probability, the two sources of information (prior knowledge and new evidence) are merged into a final estimate of probability known as Bayesian Probability (Aspinall 1992). Figure 6 displays the map generated using the Bayesian method. The resulting map perhaps is an overly simplistic implementation but it does demonstrate how prior information can effectively be incorporated into the analysis (Eastman, J.R. et al. 1993).
Finally, our decisions can incorporate the element of risk where risk is the likelihood of an inappropriate choice being made. These decisions can be based on some presubscribed level of risk tllat we are willing to assume. What risk are we willing to accept - 10%, 50% or 90% ? Obviously, this level of risk will vary from one person or group to others. From a previous example (Fig. 5) we can create a new map of area subject to flooding at say a 50% risk. This map could be cross classified with the present land area to obtain the area inundated with ourpredicted sea level rise as a result of assuming a 50% level of risk (Fig. 7). We could, also, develop a map of different levels of risk to ascertain how these different levels affect decisions on anticipatory action, for instance, the sea rise and coastal flooding for the North Slope region of Alaska.
Another necessary ingredient for success in global change research is the development of new ways of collecting data and new databases established by long term environmental monitoring effort that is not solely relegated to the natural sciences.
"Research that relies on new data that are not well understood, or simply on a slender database, is prone to err. Conclusions that are based on a small amount of data, for short period of time or for a sampled area, may seriously misrepresent what is actually occurring." (Meyer and Turner 1994: 11).
There, also, appears to be a substantial increase in new environmentally related databases; however, many of these new databases are just old ones cloaked in a new digital jacket. These databases are just more readily available now but they still have the same problems and errors as the old ones. Discretion in their use should still be the byword.
Global change research by its very nature needs an interdisciplinary approach. If this research is restricted to the viewpoint of a single discipline, either natural or social science, key factors are more likely to be overlooked which occur peripheral to the generally tunneled vision of that discipline (Kowalok 1993). And once the human component enters the equation, it becomes essential that social and natural scientist work together. This means the continual development of new paradigms and analytical tools that merge methodologies from the two scientific cultures is of paramount importance.
Careful attention must be made to scientific quality control in our research programs. The checks and balances in this type of research that are usually part of the more organized academic disciplines is often lacking.
To end we would like to quote Meyer and Turner (1994: 9).
"Bad numbers, bad analysis, and bad predictions take years to get out of print once gotten in" and appended tothis quote is "and the credibility of the research effort (field) is severely damaged too". (Lewis and Wood 1995).
Interdisciplinary research and the increased cooperation between natural and social scientists is required if success in dealing with global change problems is to be achieved.
AES. 1994. CO2/Climate Report, Canadian Climate Centre, Issue 94-1, Downsview, Ontario
Antonov, J.l . (1993. Linear Trends of Temperature at lntermediate and Deep Layers of the North Atlantic and NorthPacifc Oceans: 1957-1981, Journal of Climate, (6), pp 19281942.
Aspinall, R. 1992. An Inductive Modelling Procedure Based on Bayes' Theorem for Analysis of Patterns of Spatial Data, International Jour. of Geographical Information Systems, (6), pp 105-121.
Blaikie, P. 1985. The Political Economy of Soil Erosion in Developing Countries, Longman Scientific, NY.
Carver, S. J. 1991. Integrating Multi-Criteria Evaluation with Geographical Information Systems, Int. J. Geographical Information Systems, (5) pp 321-339.
Chapman, W.L. & J.E. Walsh. 1993. Recent Variation of Sea Ice and Air Temperature in High Latitudes, Bulletin of the American Meteorological Society, (74), pp 33-47.
CIESIN. 1992 Pathways of Understanding: The Interactions of Humanity and Global Environmental Change, Consortium for International Earth Science Information Network.
Eastman, J. R. et al. 1993. GIS and Decision Making, vol. 4, UNITAR, Geneva.
Houghton, J. T. et al. (eds). Climate Change 1992: The Supplementary Report to the IPCC Scientific Assessment, Cambridge University Press, Cambridge, UK
Kahl, J.D. et al. 1993. Tropospheric Temperature Trends in the Arctic: 1958- 1986, Jour of Geophysical Res., (98), pp 12,825-12,838.
Karl, T. 1993. Missing Pieces of the Puzzle, Research and Exploration, (9), pp 234-249.
Kowalok, M.E. 1993. Common Threads: Research Lessons from Acid Rains, Ozone Depletion, and Global Warming, Environment, (36), pp 12-20.
Kuhn, T. 1970. The Structure of Scientific Revolutions, University of Chicago Press, Chicago.
Lewis, J.E.1989. Climate Change and Its Effects on Water Resources for Canada. A Review, Canadian Water Resources Journal, (14), pp 34-55.
MacDonald, G. et al. 1993. Rapid Response of Treeline Vegetation and Lakes to Past Climate Warming, Nature, (361), pp 243-246.
Marsh, G.P. 1965. Man and Nature; Or, The Earth As Modified by Human Action, (orig. 1864) Harvard University Press, Cambridge MA.
Meredith, T.C. et al. 1994. Canadian Critical Environmental Zones: Concepts, Goals and Resources, Report 94-1, Canadian Global Change Program Technical Report, Ottawa.
Meyer, W.B. & B.L. Turner II. 1994. Cry Wolf, Human Dimensions Quarterly, (1), pp 912.
Monserud, R.A. et al. 1993. Global Vegetation Change Predicted by the Modif ed Budyko Model, Climatic Change, (25), pp 59-83.
Phillips, B. 1971. Social Research: Strategy and Tactics, The Macmillan Co., NY.
Stern, P. et al. 1992. Global Environmental Change: Understanding the Human Dimensions, National Academy Press, Washington, DC.
Thomas, W.L. Jr.1956.Man's Role in Changing the Face of the Earth, University of Chicago Press, Chicago.
Turner, B.L. II et al. 1990. Two Types of Global Environmental Change, Environment, (1), pp 14-22.
USGCRP.1995. Forum on Global Modeling, U.S. Global Change Research Program Report 95-01, Washington, DC.
Return to Natural Resources