Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Climatic and Hydrologic Systems COORDINATOR: ROBERT E. D1CK1NSON The climate system* consists of many linked components, involv- ing the atmosphere and its interactions with the oceans, land surface, cryosphere, and biosphere (see Figure 1~. Various aspects of the hy- drological cycle are important to all these components. The climate system and its manifestations in the hydrological cycle are central to the description, understanding, and prediction of the processes of global change. Human activities are capable of producing large changes in the global climate system through massive alteration of the concentra- tions of radiatively active trace gases, especially CO2, CH4, and the CFCs. The atmospheric concentrations of CO2, CH4, and other im- portant trace gases are maintained by biogeochemical cycles. Warm- ing from an increase in radiative forcing, promoted by human activ- ities, will alter the global distribution of temperature and moisture on time scales of at least decades to centuries and possibly over much A draft of this paper was prepared by committee member Robert Dickinson and revised according to comments received from a wide range of scientists (see the appendix to this paper). *The concept of the climate system was used initially by climate modelers in the early 1970s to represent this linked system. More recently, research on the physical aspects of the climate system has been organized internationally through the World Climate Research Program and within the United States through the National Climate Program Office. 107
108 THE CLIMATE SYSTEM me' Fib Atmosphere Cryosphere . FIGURE 1 The climate system. Adapted from the National Climate Program Office's National Climate Program Five Year Plan 19881992. longer time scales. Such changes will influence conditions for life over the earth. The detection and understanding of these changes are limited by an inadequate understanding of the natural variability of the system. Figure 2 (Jaeger, 1988) shows a range of possible scenarios for global warming. The figure allows for emissions of all the trace gases affecting climate and for the clelay of global temperature increase resulting from oceanic heat uptake. A narrowing in the uncertainty of the global average changes in the climate system is obviously needed. However, knowledge of such global averages alone is not very useful without an understancling of the actual change that wiB be experienced locally and regionally. A more complete picture re- quires understanding and predicting the regional manifestations of global climate change over time, with emphasis on changes in the hydrological system. The need to improve and test climate models arises from the need for better descriptions both of present and past states of the cTi- mate system and of its natural variability. Documenting the changes within the climate system, using well-validated and well-caTibrated climatic and hydrological observations, is an early goal for the global change program. New observational technologies will provide better
109 LU I 4 ~ 3 CC LU CL LL cc 1 m o o j 1 1 Upper Scenario (rate 0.8 C/decade) \~/ Middle Scenario / (rate 0.30C/decade) Low Scenario / (rate 0.06 C/decade) >I . / -1 1 1 1860 1900 1940 1980 1 1 1 1 1 1 1 1 1 1 2020 2060 2100 YEAR FIGURE 2 Scenarios of changes in globally averaged temperature that might develop in response to continued emissions of greenhouse gases. Values are plotted as differences from the 1985 value (Jaeger, 1988~. detailed descriptions and contribute to improved understanding of the complex global physical systemsthe atmospheric, oceanic, ter- restrial, and hydrologic systems including biota, snow, and ice. The exchanges of energy and water mass between these systems must be better described, quantified, understood, and predicted. Obser- vations from space have provicled, for the first time, the means to survey the many features of our planet rapidly, efficiently, and gIob- aDy with high resolution using a single instrument or coordinated group of instruments. So revealing are these sateHite-based investi- gations that they are now recognized to be indispensable for future research on the climate system. At the same time, many important aspects of the system can only be studied through extensive observations made in situ. For example, element cycling between glacial and interglacial stages over the last 2 million years is particularly well recorded in the geological record. Isolation of the physical processes that drive these massive changes is a key ingredient for understanding the climatic system in
110 its present state and hence improving our ability to predict future climatic states. Conceptual modeling and numerical simulations are and win con- tinue to be powerful surrogates for and synthesizers of global obser- vations. Elaborate data and information systems from the national level to the level of individual institutions are needed to combine information from various sources and from different federal agencies to describe the earth system as a related set of interacting processes, rather than a collection of individual components. CLIMATE FORCING FUNCTIONS The climate system changes either in response to alteration of climate forcing functions (external forcing) or as a result of Tong- time-scale internal dynamics of the system (internal forcing). With respect to the dynamics of the total earth system, the only regular external forcing terms are solar radiation and the energy released by the decay of radioactive nuclides from the earth's interior. Episodic astronomical events such as collisions with large asteroids are very rare in the earth's history, but when they occur they can have drastic effects. With respect to individual processes, it may be convenient to regard also as external forcing the radiative effects of atmospheric constituents as well as those modifications of the land surface that change so slowly that they can be considered only weakly coupled to the climate system on the decadal time scale. However, for con- sistent predictions of the (long-term) evolution of the global system, biogeochemical forcing will have to be considered as an internal part of the system. Solar and Geological Forcing Functions Solar Output and Orbital Variations The radiative energy of the sun, centered in the visible and near-infrared wavelengths, serves through its differential input as the principal driver of atmospheric circulation. Thus radiative energy powers the climate machine. The march of the seasons is ample ev- idence for the acute sensitivity of the environment of the earth to changes in the distribution of solar radiation. A more sensitive gauge of the sun's impact on climate, however, is found in the Milankovitch effect, by which changes in the earth's orbit and axial orientation
1'1 provide slow and subtle changes in the latitudinal and seasonal dis- tribution of insolation. These variations, far smaller than seasonal variations but of much longer period, serve to pace the recurrence of ice ages every 21,000, 41,000, and 100,000 years. The solar irradiance has been measured accurately only since 1979. The small decreasing trend of 0.02 percent/year reversed in 1987, apparently with the 11-year solar cycle. For the period of observed decrease, the change in radiative forcing of the earth- troposphere system was comparable with, but opposite to, that aris- ing from increasing atmospheric CO2. Solar ultraviolet radiation, varying irregularly with solar activity, dictates the basic chemistry of the upper and middle atmosphere, including the equilibrium and composition of important trace gases such as ozone. Lightning is more frequent over continents than oceans, presumably as a result of the terrestrial concentration of ionization sources linked to soils and airborne dust. The global electric field is maintained by thunderstorm activity. Solar-induced variations in cosmic ray flux or changes in the efficiency of coupling between the solar wind and the high-latitude ionosphere may perturb it, with a possible but little-studied influence on climate. Variations in solar radiation and particle fluxes have substantial effects on the magnetosphere and ionosphere, and solar irradiance variations and particle precipitation can affect the upper atmosphere. Variations in the abundance of cosmic rays alter the production rate for i4C, which is the basis for much of the dating of records of past climate over the last 50,000 years. VoIcanic Activity VoIcanic eruptions may inject into the stratosphere SO2, which can condense to aerosols and spread gIobaBy, thereby modifying ra- diative fluxes into the troposphere. Monitoring these aerosol clouds and documenting the climatic response can help the interpretation of trends in surface temperatures and contribute to an understand- ing of atmospheric transport properties. Individual large eruptions that have occurred since the beginning of the availability of global temperature records are known to have decreased global mean tem- perature by as much as 1°C for up to a few years. How great an effect might be possible from larger and/or more frequent eruptions is not known. The relatively uncommon eruption of a major caldera system could potentially have major, even if transient, influences on
112 global climate. Volcanism may provide varying amounts of CO2 over 10-m~lion-year time scales. Tectonic, Geothermal, Isostatic Rebound, Geomorphological, and Soil Changes Tectonic modifications of continental positions and shapes on 10- million-year time scales are implicated in very large climatic changes. Past variations of tectonic and volcanic factors may help test our un- derstanding of the climatic response to large forcing, both as a surro- gate verification of climate model performance and for understanding sea level changes and their potential climatic connections. The geothermal flow of heat from the earth's interior, an essen- tially constant boundary condition, supplements solar heating, albeit by only a small amount (1 part in 104~. However, this heat source is significant for the thermal regimes of permafrost and ice sheets. The response of the earth's lithosphere to loading and unloading of surface materials, including large ice sheets, sediments in deltas, and groundwater withdrawal, significantly affects regional sea levels. The response of bedrock to ice sheets has major implications for the dynamics of continental ice sheets. The geomorphological processes that move, remove, or (reposit soil and other sedimentary materials over the land surface modify the landscape and hydrological regime in significant ways. Rock weathering and soil formation processes, likewise, influence the land surface and are linked to biogeochemical cycles, especially the global carbon cycle. Wind-blown soil modulates atmospheric radiation and thus may contribute to regional climate. Continental Ice Sheets Continental ice sheets, currently those of Greenland and Antarc- tica, have large effects on climate, especially in high latitudes. They also store much of the worId's fresh water, and for at least the last 20 to 40 million years have been modifying sea level. Ice sheets normally change significantly in size only over time scales of several centuries or more. Their growth en cl decline, associated with changes in the earth's orbit and other causes, have produced past ice ages and interglacial periods.
113 Orbital Changes Changes in the earth's orbit and axial orientation modulate the latitudinal and seasonal variation of solar radiation and so apparently force a significant fraction of the glacial-interglacial fluctuation. Biogeochemical Forcing (Natural and Anthropogenic) Carbon Dioxide The concentrations of CO2 in the atmosphere have risen over the last century from 280 ppm to 350 ppm and are projected to continue to increase by 1 to 4 ppm/year over the next century. These increases in CO2 are derived from anthropogenic sources, primarily from the burning of fossil fuel and to a lesser extent from land use changes, e.g., the clearing of forests. CO2 concentration also depends on what fraction of CO2 release is taken up in the oceans (about half) and what is taken up or given off by soils and vegetation (a considerably smaller fraction) versus the fraction that remains in the atmosphere. Increasing CO2 elevates greenhouse heating by about 6 in~n/nO in W/m2 (where n = CO2 concentration, no = preindustrial values of CO2 concentration, W = watts, and m = meters). Atmospheric CO2 provi(les the carbon for growth of vegetation; thus changing CO2 modifies this growth. Changes in the chemical, biological, and physical characteristics of the oceans, vegetation, and soils in response to changing climate will provide a climate-CO2 feedback. A warmer climate may re- duce the capacity of the upper ocean to store CO2; it may also enhance both photosynthetic uptake and respiratory release of CO2. At present, we do not know the sign, let alone the magnitude of this feedback. The variations of atmospheric CO2 with glacial cycles (inferred from ciata in polar ice caps) suggest the presence of such feedbacks. Methane The concentration of CH4 has risen from preindustrial values of less than 0.S ppm to current concentrations of about 1.7 ppm, adding about 0.5 W/m2 of warming to the global climate system. Since the atmospheric lifetime of CH4 (about 10 years) is short in comparison with the time scale of changes in its concentration, shifts in its concentration reflect changes in the balance between sources
114 (mostly biological and anthropogenic) and destruction (mostly by OH radicals in the atmosphere). Past increases over the last cen- tury are ascribed largely to increases in CH4 sources associated with human activities in agriculture, forest clearing (biomass burning), and fuel exploitation, transport, and consumption. Nevertheless, the role of past and future changes in rates of CH4 destruction in the atmosphere also warrants careful scrutiny. Changes in temperature, inundation period, and inundation area wiD change CH4 flux from natural wetlands and permafrost regions. There may also be sub- stantial release of CH4 from methanehydrates present in continental slope sediments as the ocean responds to atmospheric warming. Dimethylsulfide Cloud droplets form around cloud condensation nuclei (CCN). Oceanic clouds primarily condense around sulfate aerosols, which are supplied by DMS emitted by the oceans (CharIson et al., 1987~. DMS is generated by certain kinds of marine phytoplankton, and its generation rate may depend on the temperature of the ocean surface waters. An increase in sulfate CCNs would increase the numbers of cloud droplets and hence increase planetary albedo over the oceans. We need to better quantify how the flux of DMS now influences cloud cover and albedo and how various changes might modify the flux of DMS to the atmosphere. Other Aerosols Aerosols of both natural and anthropogenic origin, e.g., from desert dust or conversion of combustion products, may modify radia- tive fluxes either directly by their radiative properties or indirectly through their effects on cloud cover or optical properties. Lifetimes of aerosols depend on wet and dry deposition processes, which may vary with climate change. Other Trace Gases The CFCs Fell and F-12 may in several decades also build up to large enough concentrations to add significantly to greenhouse warming. Likewise, N2O, tropospheric and stratospheric 03, and to a lesser extent many other trace gases are also of radiative impor- tance and are undergoing changes in flux rates. Changes in climate and hy~lrology will feed back on production and destruction rates of
115 these trace gases. In particular, O3 decrease and CO2 increase can reduce stratospheric temperatures by large amounts and so affect the planetary wave and radiative coupling to the troposphere. Strato- spheric water vapor is strongly dependent on the temperature of the tropical tropopause and on CH4 concentrations. Effects of [and Use Changes on Climate Changes in agricultural activities, deforestation, and desertifi- cation are increasingly affecting the climate system and in turn are affected by changing climate. The impacts of vegetation and land use changes on climate are of two kinds: 1. Biogeochemical: surface modifications affect global atmo- spheric composition and its radiative balance. Biomass burning, a major mode of deforestation, is a source of CO2, CH4, N2O, CO, particulates, and other trace gases. Deforestation is generally ac- companied by degradation of soils and enhanced fluxes of CO2 and N2O to the atmosphere. Increases in rice cultivation and ruminant population increase the amounts of CH4, while an increase in fertilizer usage contributes to rising N20 concentrations in the atmosphere. 2. Biogeophysical: modifications of vegetation cover alter re- gional hydrology and regional surface-energy balance. The impor- tance of land-surface effects on climate has been suggested by model- ing sensitivity studies. These studies indicate that increases in albedo from land degradation in semiarid regions could promote drought. Other studies have shown that model-simulated climates are also sensitive to land-surface boundary conditions that affect evapotran- spiration. Interactions between vegetation and snow cover may also be important. The issue of realistically modeling how land processes, in partic- ular vegetation and soil moisture, interact with the climate system is only now beginning to be addressed. Such modeling should recog- nize the two-way coupling between these two systems, as vegetation changes both its form and its function in response to climatic forc- ing. The modeling efforts must recognize the small spatial scales over which vegetation cover and soils vary.
116 RESPONSE OF THE CLIMATE SYSTEM TO CHANGES IN FORCING Nature of Climatic Forcing Changes in total solar output, stratospheric aerosol, or atmo- spheric trace gases that are Tong-lived (fairly well mixed gIobally), such as CO2 and CH4, affect the climate system through the addi- tional global radiative forcing that they provide. Some differences occur in the vertical, latitudinal, and seasonal distribution of atmo- spheric radiative forcing, but the effects of these differences are not yet established and are apparently relatively small. Thus studies of the global response of climate to increases in atmospheric CO2 also largely apply to increases of other trace gases and to increased so- lar heating. The practical question is thus how the climate system responds to the sum of various global inputs in the past and future. The change in climate from changes in global radiative forc- ing will be in part manifested by a variety of regional responses, with changes in some regions more significant than those in others. Changes in other climate forcing factors, in particular those related to vegetation and ice cover, win have their largest effects on a re- gional scale and be specific to the given forcing and location. These effects largely involve changes in energy exchange processes at the surface, especially over continental interiors. Key Areas of Uncertainty in Evaluating Climate System Response Cloud Effects Clouds have multiple properties that are important for climate change. They modulate both incoming solar radiation and outgoing thermal infrared radiation. Reflection of solar radiation depends not only on cloud amount but also on the optical thickness of clouds, on the liquid water content of the clouds, and on the size distribution of the cloud (lroplets. The upward flux of thermal infrared emission from clouds depends on cloucI-top temperature and on cloud emis- sivity. Only high, thin clouds, i.e., cirrus, have emissivities that are significantly less than 1.0 and hence have thermal emission controlled by cloud thickness. Preliminary and incomplete treatments of cloud feedback on cTi- mate have been included in recent GCM studies of climate response
117 to CO2 doubling. These studies have found that cloud changes sig- nificantly affect the surface temperature response. As temperatures increase, clouds are expected to hold more liquid water ant! hence be more reflective, thereby reducing the warming significantly. In the case of cirrus, however, increased amounts of thermal infrared radiation are expected to be trapped and so amplify the greenhouse warming. Unfortunately, these effects have not yet been included in GCM simulations of climate change. Different GCM simulations have shown qualitatively similar spatial patterns in the TongitudinaDy averaged cloudiness change; they indicate, in particular, a decrease in cloudiness in the moist, connectively active regions such as the tropical and midcIle latitude rain belt and an increase in the stable region near the surface from micIdle to high latitudes as wed as in the Tower stratosphere. Such cloud changes would be expected to have significant regional effects on radiative balance and hence surface temperature. Actual clouds are often thinner than a model layer and generally have horizontal scales of 1 to 100 km, i.e., scales unresolved by the grid of global models; hence, their radiative properties may not be correctly modeled. Thus we need appropriate data bases and a conceptual and statistical framework for describing the morphology of realistic cloud fields made up of clouds of various shapes and sizes. Also required is an understanding of the radiative, dynamical, and microphysical processes determining this morphology. Diurnal variations of cloudiness are poorly characterized but are probably significant for determining changes in cloud radiative balances and land surface evapotranspiration feedbacks. The net global feedback is even more difficult to mode} than the regional effects of cloud changes. Cloud processes in models need to be related through observations to other meteorological processes. The primary line of evidence that actual cloud feedbacks are not extremely far from what is modeled is the reasonable success of global models in reproducing seasonal and diurnal cycles of the global climate system. High-Latitude Response The largest regional temperature increases from global warming are expected in high latitudes as a result of albedo and atmospheric stability changes linked particularly to the extent of sea ice. Modeling
118 studies find that most of the high-latitude temperature increase oc- curs in winter as a result of increased summertime uptake of heat by the oceans through an increased fraction of open water and thinner sea ice and a corresponding wintertime release of that heat. The annual growth and decay of sea ice introduce a strong sea- sonal signal not only to the physical but also to the biological com- ponents of the polar regime and may affect ocean uptake of CO2 and nutrient supplies. Treatments of these issues have been handicapped by inadequate consideration of the complex effects of sea ice dynam- ics, ice distributions at subgrid scale, and interactions with ocean dynamics and salinity and with overlying cloud cover. Also, polar re- gions, especially in winter, are relatively poorly represented by global climate models because of their low absolute concentrations of water vapor and the singular aspects of the poles in spherical coordinate systems. Processes determining albedos of snow and ice are linked to many other components of the system. Changes in permafrost have large effects on the land surface and hydrological regimes. Increases in summer temperature and thaw period may enhance photosynthesis and respiration of boreal forests as well as CH4 production by high- latitude peat lands. Atmospheric Hydrological Response Water vapor is the most variable radiatively active constituent of the atmosphere and is the main contributor to the greenhouse effect of the atmosphere. In addition, even slight changes in the water mix~ng- ratio and corresponding changes in relative humidity can alter the density, optical properties, and altitude of clouds and hence affect the transfer of solar and terrestrial radiation through the troposphere. A rise in global tropospheric temperature increases atmospheric water vapor, which in turn increases by approximately twofold the rise in surface temperature from the other greenhouse gases alone. However, this feedback is rather uncertain because of lack of understanding of possible changes in water vapor concentrations, especially in the upper troposphere and lower stratosphere. Global circulation models simulate the very broad-scale features of global precipitation reasonably wed, including the main bands of precipitation along major depression tracks and in the tropical rain belts and the areas of meager rainfall over major deserts and along the eastern side of the subtropical oceans. Nevertheless, there are
119 serious errors on the regional scale, and patterns and amounts of precipitation vary from model to model. Even the degree of error is difficult to jucige because of major uncertainties in measurements. In the long term, the radiative cooling of the atmosphere is bal- anced globally by latent heat release resulting from net condensation and the transfer of sensible heat from the surface. Thus simulated precipitation rates are sensitive to the parameterization not only of precipitation processes but also of radiative and turbulent transfer of heat from the underlying surface. At the low horizontal resolution of current global models, the simulations of sharp features such as the tropical rain belts and frontal systems are smoothed. The orographic effects of mountains and ice sheets are poorly treated. The high spatial variability of rainfall patterns, especially those connected with convective systems, presents serious challenges both to estimation of rainfall amounts and to the modeling of variations of the surface-hydrologic system over land. Statistical representations of rainfall structures as now used in studies of catchment hydrology need to be related to atmospheric processes and incorporates! into climate models. Imp roved algorithms for simulating the movement of water va- por, sensible heat, and momentum through the planetary boundary layer are important for modeling the coupling between the atmo- sphere and the land and ocean surfaces. Surface Hydrological Response The terrestrial hydrology is forced by atmospheric phenomena, i.e., precipitation, surface radiative balance, surface winds, air tem- perature, and humidity. The land in turn interacts with the atmo- sphere through fluxes of sensible and latent heat. Key constraints on these interactions are the conservation of energy and water, and these constraints must be included in even the simplest treatments. However, actual land surfaces are extremely complex, and models are only now being developed that capture some part of this complexity. Surface hy(lrological processes, i.e., runoff, infiltration, and evap- oration, (lepend on the soil, vegetation, and the aspect and slope of the land surface. These processes appear to vary over a wide range of spatial scales. The parameters entering into the descriptions of these variables in models are inferred from local measurements of processes made in specific environmental settings. Precipitation is initially intercepted by canopy foliage; that which reaches the ground
120 percolates downward, according to soil-and moisture-dependent hy- draulic conductivities. Water is returned to the atmosphere either by evaporation from the soil surface or by evapotranspiration from the surfaces of green leaves. The movement of water through plants depends on plant phys- iology, in particular leaf stomata! functioning, which is not always easily modeled. These physiological processes interact with atmo- spheric gases. For example, CO2 concentrations are known from laboratory studies to have a large immediate effect on the stomata! resistance of many plants. However, adjustments that may occur over the lifetime of plants or in future generations that may compensate for such effects are poorly known. Roots that move water from soil to plants can provide the limiting resistance to evapotranspiration under conditions of water stress. The rate of evapotranspiration is proportional to differences be- tween the water vapor concentrations at or just within soil and plant surfaces and the water vapor in the overlying air. The vertical struc- ture of vegetation coupled with the complexities of turbulent transfer processes within and immediately above the canopies complicates simple flux-gradient modeling. The vapor gradients driving fluxes of water from soil and vegetation surfaces depend in turn on surface temperatures, which themselves depend crucially on the processes determining the surface energy balance. Climate models that are used to simulate future climate change from increasing trace gases have not yet incorporated detailed mod- els of hydrological processes. However, simple energy- and moisture- conserving parameterizations for soil moisture have resulted in an indication of possible future hydrological change. In particular, some model results suggest a significant decrease in soil moisture in mid- dle latitudes of the Northern Hemisphere during the summer season. This summer drying could arise from a northward shift in the mid- latitude summer rain belt and from a greater and more intensive period of drying associated with earlier snowmelt and warmer tem- peratures. Decreases in cloud cover may amplify this m~dcontinental drying, raising the specter of a possible future increase in the fre- quency and severity of agricultural drought in central continental areas of the mid-latitudes. Forest fires and dust storms also must be consi~lerec3 as potential agents of land surface change. The water supply for mid-latitude arid regions, mainly derived from mountain winter snowpack, is poorly represented in current models of climate change, where resolution is too coarse to resolve the topography of
121 indiviclual mountain ranges and their critical effects on precipitation. Treatments of runoff in climate models do not adequately represent the realities of catchment-scaTe hydrology, such as the highly non- linear, almost threshold, relationships between rainfall and runoff as wed as the dependence on terrain, soil heterogeneity, and subgrid- scale statistical structures of rainfall patterns. The hydraulic con- ductivity of soils, determining the rate of water infiltration, can be strongly controlled by biological processes that influence, for exam- ple, soil structure and macropores. Improvements in modeling the linkages between surface hydrol- ogy and climate require better understanding of evapotranspiration from diverse terrestrial vegetation, other land covers, and lakes; in- fiItration of water through the unsaturated zone with consequent improved understanding of aquifer recharge; the conversion of pre- cipitation to streamflow, including the role of topography at a variety of scales of catchment and storm patterns; and the processes of snow accumulation, transformation, redistribution, and melting. Ocean Coupling to the Rest of the Climate System The ocean is coupled to the atmosphere through exchanges of sensible, latent, and radiative heat; momentum; various important trace gases, e.g., CO2 and DMS; and the exchange of water by precip- itation evaporation and river flow from the land. The temperature of the ocean surface is especially important in determining these fluxes. Surface temperatures in turn depend not only on these fluxes but also on the redistribution of thermal energy within the oceans by vertical mixing and by horizontal energy transport, i.e., the oceanic general circulation. The ocean circulation is in turn sensitive to the supply of fresh water and consequent changes in salinity. Oceanic heat uptake will affect the transient evolution of both global temperature increase and changes in regional hydrological properties. In high latitudes, the presence and seasonal variation of sea ice constitute another important oceanic climate factor that couples to atmospheric and oceanic energy exchange processes and can substantially enhance the response of high-latitude climate to radiative forcing. The ocean, through its intermediate-time-scale circulations, is capable of changing the equilibrium concentrations of CO2 on rela- tively short time scales. The possibility of multiple circulation states of the ocean leads to the possibility of multiple states of the climate
122 system. Exchanges of nutrients between land and ocean and oceanic nutrient cycles may also have climate feedbacks. Abrupt Change ObservationaBy, an abrupt change is one that is rapid in relation to the time and space resolution of direct or proxy measures of change. Considerations of past and future climate change have been dominated by the assumption that any gradual forcing would elicit a smooth response. However, an increased understanding of the global climate system and its complex feedback processes suggests the possibility of changes that come in steps rather than occur gradually. For example, records from ice and ocean cores, and conceptual and numerical models, suggest that the Atiantic Ocean could have more than one mode of circulation. Abrupt transitions between these modes might be triggered by variations in salinity of the high-latitude oceans caused by changes in river flow, glacial melt water, sea level, precipitation, or evaporation. Other features of oceanic circulation, which in turn strongly affect the rest of the system, may shift abruptly in response to slow shifts in atmospheric circulation and rainfall patterns. Large increases in the frequency or intensity of E! Nino Southern Oscillation events in the tropical Pacific might occur. These events have a major influence on rainfall patterns, especially in th tropics, and on oceanic circulations that determine nutrients for marine productivity. Disappearance of year-round arctic ice cover because of greenhouse warming might be accompanied by significant changes in high-latitude climate. The large lake systems that existed in the Great Basin of the western United States about 13,000 years ago almost completely disappeared within a few hundred years, suggesting that a rapid n.nc3 discontinuous change occurred in the overall behavior of the surface hydrologic system for a region covering about one-half million square kilometers. Other, perhaps speculative, abrupt changes include the collapse of the west antarctic ice sheet, catastrophic CH4 release from destabilization of CH4 clathrates in the Arctic, a rapid destruction of the Amazon forest regional climate system, and a discontinuous change to a colder and less disturbed stratosphere in the Northern Hemisphere that is accompanied by a large decrease in high-latitude stratospheric 03. Where there are large gradients, e.g., in the tropical rainfall distribution, climatic shifts imply large changes in regional climates.
123 Some or all of these possibilities appear rather unlikely over the next century, but their consequences could be much more severe than those of gradual global warming. Thus there is a clear need for scrutiny of evidence in the climatic record for abrupt climate change and its effects on the system from natural causes. Careful observation of the current system and studies with joint ocean-atmosphere models of the climate system that may exhibit multiple equilibria are needed to better understand the prospects of natural and human-induced abrupt change. The links between planetary radiation, atmospheric trace gas content, atmospheric water transport, ocean circulation, and the marine and terrestrial biospheres also need to be examined with this possibility in mind. D O CUMENTATION OF PRES ENT CLIMATE AND THE FACTORS THAT CAUSE IT TO CHANGE The following key research issues should be addressed: i 1. The recovery of the past history of environmental change, evolving studies of ocean and lake sediments, ice cores, tree rings, and sea level changes for information about past climate, hydrolog- ical regimes, ocean circulation, biological interaction, atmospheric chemical composition, and solar inputs. 2. Comprehensive documentation of changes in the current phys- ical environment, including observation and understanding of the sun and near-earth space, the atmosphere, snow and ice, oceans, and the sails and vegetation of the earth. The former aspect is discussed in the companion paper on "Earth System History and Modeling," and the latter is discussed below. Only within the last few decades have the technology, telecom- munications, and the logistics necessary to handle large amounts of data been developed to the point where comprehensive global ob- servations of the climate system have become feasible. Sustained progress in understanding the global system wiD require a compre- hensive continuous observing program, ongoing theoretical investi- gations, and global modeling efforts. Such a program is necessary to document climate changes, to further studies of important processes, and to provide the data needed to construct, test, and utilize models of the system and its components. Some limited instrumental data extend back over a century or more and provide much of what is now known about past change. Such data must be analyzed and their
124 collection extended into the future, with considerable care for the continuity and calibration of these records. Most studies of the current global atmosphere now rely on the information provided by the worldwide network of surface and upper- air sounding stations organized by the World Weather Watch and the system of four to five geostationary and two polar-orbiting me- teorological satellites, which has been, more or less, in operation since the Global Weather Experiment of CARP (about 1979~. This real-time operational observing and data management system is com- plemented by various sources of delayed data such as temperature profiles of the upper oceanic layer, sea level data, and a wide variety of surface and subsurface hydrological measurements. The system suffers from several deficiencies, which include (1) serious gaps in the three-dimensional distribution of wind observations over the oceans; (2) large uncertainties, resulting from insufficient sampling and mea- surement errors, in the estimation of precipitation over land areas and total lack of information over the oceans (and over some coun- tries that do not make hydrological data available for international exchange); and (3) significant uncertainties in the three-dimensional global distribution of water vapor and clouds, and the almost total lack of global information about surface properties, including soil moisture. Climate studies are concerned with the quality and con- tinuous availability of the data, whereas the real-time collection is primarily of importance for operational agencies. Current prospects for the geostationary platforms of the system are satisfactory, since the major meteorological satellite operators are now in the process of finalizing their plans to replace some of the existing operational satellites with a second generation of space- craft with nearly identical advanced sensors. However, the replace- ment of existing operational environmental satellites in polar orbit must be undertaken to ensure the continuity of essential atmospheric measurements as well as to provide adequate facilities for testing experimental remote sensing instruments. The measurement of the distribution of all phases of water in the atmosphere should be given high priority if the fundamental role of fresh water on time scales of decades to centuries in ah subsys- tems of the earth system is to be understood. Global patterns and amounts of precipitation on a year-to-year basis must now be es- timate(1 by the use of rain gauge networks over land coupled with correlations of precipitation with cloud top temperatures over oceans. A Tropical Rainfall Measurement Mission would begin to satisfy a
125 critically important requirement for direct space measurements of global precipitation by testing the feasibility of using active and pas- sive microwave data together with visible and infrared imagery. A Tow-incTination orbit could resolve the mean diurnal cycle of rainfall over the tropics and assess the relationship of latent heat released into the atmosphere to anomalies in atmospheric circulation. Studies of atmospheric water vapor are also equally important. Its variations in space and time are inadequately known. Planned improvements of operational satellite instruments should start soon to yield significant information for the lower troposphere on a global scale. Atmospheric radiation balance is especially sensitive to the concentrations of water vapor in the upper troposphere and Tower stratosphere. Radiative balance of the climate system, and the dynamical pro- cesses that give rise to cloudiness, are highly significant for climate. Continuous measurements of the total solar output provide a funda- mental boundary condition for radiative inputs, as do measurements of the varying contribution to planetary albedo from stratospheric aerosols. Current analyses of cloud amounts and their trends are based upon visual estimates by ground-based observers. The Inter- national Satellite Cloud Climatology Project (ISCCP) now under way will assemble a 5-year data set of radiance measurements from information returned by five geostationary and two polar-orbiting satellites. The data wiD allow compilation of meaningful statistics for cloucT amount, type, height, and optical thickness. The First TSCCP Regional Experiment (FIRE) is a U.S. contribution helping to vaTiciate algorithms for this derivation. Global information on the altitude of cloud bases is not included in these measurements, but is also needed. A complementary approach to determining the role of clouds in the radiation balance is the measurement of the components of the radiation balance itself: the fluxes of solar and infrared ra- diation at the top of the atmosphere and their inferred values in an equivalent cloud-free environment. The difference is the "cloud forcing." This quantity is obtained using a combination of a sim- ple, wide-field-of-view instrument on sun-synchronous polar-orbiting satellites together with non-sun-synchronous satellite measurements of the Earth Radiation Budget Experiment (ERBE). Spectral and directional models are necessary to relate the narrow-spectral-band, narrow-field-of-view operational instruments to the total fluxes and
126 to infer the fluxes at the top of the atmosphere from those at satellite altitude. The International Satellite Land Surface Climatology Project (TSI`SCP) program, through such field programs as the Hydrologi- cal Atmospheric Pilot Experiment (HAPEX) in 1986 and the First ISESCP Field Experiment (FIFE) in 1987, focuses on the develop- ment and improvement of algorithms for remote sensing of physical properties of the land surface important for climate models. Progress has been made in the observation and interpretation of a vegetation index based on the difference in reflected radiances in the visible and near-infrared channels of the Advanced Very High Resolution Ra- diometer (AVHRR) imager and, more recently, in another approach to a vegetation index based on a microwave measurement. Also promising are procedures to measure surface radiative temperature, including its diurnal variations, from window infrared and microwave thermal emission and procedures to obtain incident surface radiation from reflectance of visible radiation by clouds. Progress in inferring surface albedo is also being made, with improvement in atmospheric corrections including clouc] removal, better calibration, and advances in unclerstanding the angular dependence of bidirectional reflectance of lancl surfaces. Observation from space of land surface climate properties re- quires considerable information about the atmosphere. First, radia- tive emissions sensed from the surface are modulated by atmospheric gases, clouds, and aerosols, whose effects must be quantified. These modulating influences can be as important as the land properties being measured, both for the dynamics of the land processes and for the climate system in general. Second, some key observations such as the indirect measurement of soil moisture, hence evapotranspiration, require measurements of near-surface atmospheric properties, e.g., humidity and winds. A detailed description of surface-energy bal- ance over land, like that over the ocean, requires both atmospheric and surface information. Observational research on land surface climate processes and hydrology must be closely coordinated with related meteorological studies, such as those of the U.S. National STORM Program, and with global change activities in the areas of biological dynamics and biogeochemical cycles. Besides satellite systems, there will be a need for continued application and development of aircraft measurements, in situ sounding systems, and surface instrumental systems, espe- ciaDy for process studies. These land surface climate processes are
127 [Linked to the many practical facets of the impacts of climate change, e.g., water resources and agricultural productivity. Changes in glaciers and small ice caps, in both high-latitude and mountain environments, contribute to global sea level rise and fall. Seasonal snow is another sensitive indicator of climate change and provides positive feedback through its effect on global albedo. The presence of sea ice, already measured from space, completely changes the magnitude of the heat and moisture fluxes from the ocean as weD as the surface albedo. High spatial resolution of measurements is important to determine the fractional ice-covered areas in regions of broken ice. Large thermal fluxes can be associated with these marginal ice zones. Additional measurements are needed of sea ice thickness, the distribution of ice of differing ages and its motion, and the extent of surface melting and its relation to ice albedo. A change in the volume of polar ice could result from an increase or decrease in global temperatures. Yet at present we have no reliable means to assess changes in the inventory of ice in the Greenland and antarctic ice sheets. We need to know whether present ice sheets, averaged over the seasonal cycle, are in steady state, growth, or decay phases. Variations in polar ice volume should be detectable in sea level measurements, but these data are noisy and influenced by the thermal expansion that results from global temperature change, by local tectonic movements, and by rebound from the melting of the last major glacial ice sheet. Ground-based surveys are not definitive, but satellite altimetry could be applied to this problem. Current observational programs in the ocean sciences support the goals of the IGBP to the extent that they focus on ocean sur- face temperatures and how they might vary with climate change, and on how the oceans store and exchange radiatively important trace gases. Surface temperatures are linked to vertical mixing and convection processes and depend on vertical distributions of temper- ature and salinity, horizontal advection of these quantities, and hence ultimately on the oceanic general circulation. Complementing the sparse data coverage by the many ship- based programs, satellites provide regular global observations over the ocean of surface temperatures, sea ice cover, and in the future, wind stress from roughness, winds over the ocean, and sea level height. It is necessary to measure the geoid more accurately in order to use sea level heights to derive ocean surface currents. Long-term monitoring of sea level height is needed both from surface and from space platforms.
128 PRINCIPAL ISSUES AND RESEARCH CHALLENGES Sustained, Calibrated, [ong-term Measurements Long-term monitoring of the more important global climate and hydrological variables is crucial. This objective requires stable, we0- calibrated measurements over a multidecadal time frame. Monitoring of both forcing functions and system responses is needed. The list of Tong-term monitoring needs cited in NASA's Earth System Science Committee (1988) report (here reproduced as Table I) merits fun support. The most critical forcing variables are solar irradiance, vol- canic emissions, and radiatively important trace species, especially CO2, CH4, and the CFMs. Atmospheric response variables of high- est priority for long-term measurement are surface air temperature, tropospheric temperature, precipitation, and surface pressure. Also very important are winds, especially in the tropics, components of the earth radiation budget, cloud amount, type, height, and optical thickness. Land surface properties for a variety of representative surfaces must also be measured over a long term, especially those proper- ties controlling the fluxes of water between surface and atmosphere. These properties include precipitation, measures of soil moisture, and vegetation cover, all on a regional scale. Measurements must include the surface radiative temperature, incident solar flux, seasonal snow cover, snow water equivalent, changes in the volume of high-altitude and continental ice sheets, amount of river runoff, distribution of permafrost, levels and freezing dates of lakes, extent and seasonaTity of wetlands, and other surface characteristics such as albedo, rough- ness, and emissivities in the infrared and microwave bands. Key ocean variables are sea surface temperature; sea level; sea ice extent, type, and motion; ocean wind stress; subsurface circulation; and incident solar flux. Adequate attention should he given to sustained, calibrated, long-term measurement of the types of variables diiscussed above and analyzed by NASA 's Earth System Science Committee. Information Systems Components of information systems inclu(le data transmission, quality control, directories, catalogs, and inventories; products of special analyses; status of data observation, collection, archiving, and distribution; and agreements for international exchange of data.
129 TABLE 1 Sustained, Long-term Measurements of Global Variables Important for the Study of Global Change on Time Scales of Decades to Centuries (NASA's Earth System Science Committee, 1988) Analysis Product Variable Importance Quality External Forcing: Solar ~rrad~ance *** A Ultraviolet flux ** B Index of volcanic emissions * D Radiatively and Chemically Important Trace Species: COO *** A N,O ** A CH4 ** ~ Chlorofluoromethanes ** A Tropospheric O3 ** C- CO ** D- Stratospher~c O3 *** C Stratospheric H?O ** C Stratospheric NO? * C Stratospheric HNO3 * C Stratospheric HC~ * C- Stratospheric aerosols ** B Atmospheric Response Variables: Surface air temperature *** B Tropospheric temperature *** B Stratospheric temperature ** C Pressure (surface) *** A- Tropical winds ** C- Extratropical winds * B Tropospheric water vapor ** D Precipitation *** C- Components of Earth radiation ** B budget Cloud amount. type, height ** D Tropospheric aerosols * D Variable Land-Surface Properties: Surface characteristics (for albedo. roughness, infrared and microwave emittance) Index of land-use changes (broad classification of vegetation types) Index of vegetation cover Index of surface wetness Soil moisture Biome extent, productivity, and nutrient cycling Analysis Product Importance Quality * C- ** F *** D * F *** F *** F Ocean Variables: Sea-surface temperature Seance extent Sea-ice type Sea-ice motion Ocean wind stress Sea level Incident solar flux Subsurface circulation Ocean chlorophyll Biogeochemical fluxes Ocean CO,, B D C- C D D C- C- C C Land-Surtace Properties: Surface radiating temperature * F Incident solar flux * C- Snow cover * C Snow water equivalent * F River runoff (volume) * B River runoff (sediment loading) * River runoff (chemical constituents) * F KEY Importance (for documenting and understanding global change): *** Essential ** High * Substantial Analysis Product Quality (presently available multiyear global analyses): A = Good quantitative, well calibrated B = Well discriminated, absolute' accuracy doubtful C = Useful, poor discrimination D = Qualitative index, interpretation doubtful F = No information - = Not global coverage Policies for pricing of data and funding for acquisition must encour- age, rather than hinder, the required research and analysis with Tong-term multidecadal data bases. Calibration and long-term sta- bility of operational measurements must be adequate to ensure the integrity of the long-term data bases. Plans must include adequate programs for reanalysis of model-assimilated data bases. Special attention is needed to develop systems to handle the large data streams from present satellite observations and the even larger ones from future coordinated packages as planned by NASA in its EOS program. The analysis of these data into forms readily
130 manageable by the scientific community will be a key issue. Large amounts of information must be made available to the scientific community in a friendly, supportive, and timely fashion, through distributed systems using modern computational (supercomputer) and workstation technologies. An aggressive and supportive new approach at the national level is needed to develop a national data and information system for climatic and hydrological data that is consistent among agencies and thus allows ready access by a wide variety of researchers. Hydrological Cycle Understanding the cycling of water in its three phases is crucial for studies of global change. Within the atmosphere and at the sur- face, water absorbs and reflects solar radiation with a wide range of albedos. Both clouds and water vapor are more important for trans- fer of thermal infrared radiation than any of the trace greenhouse gases. Evaporation removes much of the net radiative heating at the earth's surface; vertical and horizontal transport of water vapor is a dominant mechanism for redistribution of energy within the atmo- sphere. Water is a key ingredient of tropospheric and stratospheric chemical processes. It is also critical for the presence of life. In par- ticular the dynamics of terrestrial vegetation, through its dependence on water supply from precipitation, are linked to climate. Current information on the global and regional budgets of water is inadequate, and key ingredients i.e., clouds, atmospheric water vapor, soil moisture, precipitation, and evaporation are now inad- equately measured. These variables and the processes responsible for them need to be better represented in global models. The World Climate Research Program (WCRP) has (leveloped a concept of a Global Energy and Water Experiment (GEWEX) to greatly improve the observational and modeling basis for including the physical hy- drological system in studies of global change. Building upon the GEWEX concept of WCRP, a program should be developed to better define the hydrological cycle and! related energy fluxes by means of global measurements of observable atmospheric and surface proper- ties, and to study and mode! processes of the gloloal; hydrological cycle and its connections to land processes and properties of the oceanic surface layers.
131 Effects of Terrestrial Vegetation on Climate and the Hydrological Cycle There are close connections between climate and vegetation cover of both natural and managed ecosystems. The global patterns of net radiation, ocean temperatures, and precipitation impose strong constraints on the dynamics of vegetation. However, there are also significant feedbacks of the vegetation cover on the climate system. Important factors controlled by vegetation include surface albedo, evapotranspiration, and surface roughness. These factors are still poorly represented in global climate models. They depend in part on the changes of vegetation cover with seasonal temperatures and associated extreme weather events, and on variations in precipitation processes, seasonal or otherwise. They also depend on atmospheric composition either as a stress factor or as a source of nutrients, e.g., the dependence of evapotranspiration on atmospheric CO2 either clirectly through its control of stomata! closure or indirectly through its control of ecosystem structure. A program of observations and; modeling should be developed to improve understanding of the effects of terrestrial vegetation on climate and to build a foundation for incorporation of these effects in models of global climate change. Sources of Biogenic Gases and Dependence on Climate Some biological processes contribute to greenhouse warming by changing atmospheric composition. Other processes provide con- densation nuclei for clouds and so may contribute to art increased cloud albedo, hence a cooler climate. CH4, the next most important greenhouse gas after CO2, is poorly understood as afactorin global change. We do not know why it has increased by more than a factor of 5 since the last ice age, and we cannot predict how either terrestrial sources or atmospheric Toss might be modified by changes in climate or atmospheric composition. DMS is apparently the major source of cloud condensation nuclei over the oceans; yet we have no idea as to how its sources might change with climate change. A program of observation and research is needed to improve understanding of the sources of atmospheric gases from biological processes and to develop parameterizations for including these sources in coupled models of global climate change and biogeochemical cycles.
132 Atmospheric Dust and Aerosol Solid particulates in the atmosphere affect atmospheric radiation balance directly through their modulations of atmospheric radiation balance and indirectly through their role as cloud condensation nu- clei. Atmospheric aerosols either increase or decrease the net ab- sorption of solar radiation, depending on their optical properties and those of the underlying surface. Carbon particles are generally the most absorbing, and pure sulfate particles the most reflective. Wind-suspended material originating in arid regions is widespread and has been recorded in vastly increased amounts in the paleocTi- matic record. Its representation in climate models would help focus on the transport characteristics of the models (also needed for cou- pled chemistry studies) and would add representation of a significant ingredient in the climate system. Research is needed to describe the surface processes responsible for the lifting of soil and desert aerosol on a global [oasis, the distribution of this aerosol within the atmosphere, its global transport and locations of deposition, and the inclusion of all these processes within a global climate model. BIB [IO GRAPHY Charlson, K. S., J. E. Lovelock, M. O. Andreae, and S. G. Warren. 1987. Oceanic phytoplankton, atmospheric sulfur, cloud albedo and climate. Nature 326:655-661. Earth System Science Committee. 1988. Earth System Science, A Closer View. Washington, D.C.: National Aeronautics and Space Administration. International Council of Scientific Unions. 1986. The International Geosphere-Biosphere Program: A Study of Global Change. Final report of the Ad Hoc Planing Group. ICSU Twenty-first General Assembly, Berne, Switzerland, September 14-19, 1986. Jaeger, J., 1988. Developing policies for responding to climatic change. World Climate Program Impact Studies. WMO/TD No. 225. Geneva: World Meteorological Organization. National Climate Program Office. 1988. National Climate Program Five Year Plan 1988-1992. Washington, D.C. National Research Council. 1986. Global Change in the Geosphere-Biosphere: Initial Priorities for an IGBP. Washington, D.C.: National Academy Press. Office for Interdisciplinary Earth Studies. 1986. Climate-Vegetation Interactions, C. Rosenzweig and R. Dickinson, eds. Boulder, Colo. Office for Interdisciplinary Earth Studies. 1987. Arctic Interactions Recommendations for a Multidisciplinary Arctic Component with IGBP. Boulder, Colo. Risser, P. (ed.~. 1985. Spatial and Temporal Variability of Biospheric and Geospheric Processes. Paris: ICSU Press. World Climate Research Program. 1987. Report of the Workshop on Space Systems Possibilities for a Global Energy and Water Cycle Experiment. World Climate Program-137.
133 World Climate Research Progrmn. 1988. Global Energy and Water Cycle Experiment: Concept and Rationale. Draft report of the Joint Scientific Committee Study Group on Global Energy and Water Cycle Experiment. WCRP-1572. APPENDIX: INDIVIDUALS WHO PROVIDED COMMENTS ON BACKGROUND PAPER ON CLIMATIC AND HYDROLOGIC SYSTEMS Richard A. Anthes, National Center for Atmospheric Research D. James Baker, Jr., Joint Oceanographic Institutions, Inc. Eric ]. Barron, Pennsylvania State University Roger Barry, University of Colorado Larry Benson, U.S. Geological Survey Ralph Cicerone, National Center for Atmospheric Research William C. Clark, Harvard University John A. Dutton, Pennsylvania State University Tnez Fung, Goddard Institute for Space Science Peter H. Gleick, University of California, Berkeley Robert Gurney, National Aeronautics and Space Administration J. Michael Hah, National Oceanic ant! Atmospheric Administration Frederick Koomanoff, U.S. Department of Energy Research Arthur H. Lachenbruch, U.S. Geological Survey Louis Lanzerotti, AT&T Bell Laboratories Conway Leovy, University of Washington Jerry MahIman, Geophysics Fluid Dynamics Laboratory G. A. McBean, Institute of Ocean Sciences Mark F. Meier, University of Colorado, Boulder PhiDip Merilees, National Center for Atmospheric Research Ellen S. MosTey-Thompson, Ohio State University Marshall E. Moss, U.S. Geological Survey K. T. Paw U. University of California, Davis Veerabhadran Ramanathan, University of Chicago CIaes Rooth, University of Miami Jorge Sarmiento, Geophysics Fluid Dynamics Laboratory Stephen Schneider, National Center for Atmospheric Research Piers SeDers, NASA Goddard Space Flight Center Kevin Trenberth, National Center for Atmospheric Research Tom Van der Haar, Colorado State University J. M. Wallace, University of Washington Warren Washington, National Center for Atmospheric Research