Biospheric Aspects of the Hydrological Cycle

Journal of Hydrology 212–213 (1998) 1–21 Preface Biospheric Aspects of the Hydrological Cycle R.W.A. Hutjes a,*, P. Kabat a, S.W. Running b, W.J. Shuttleworth c, C. Field d, B. Bass e, M.A.F. da Silva Dias f, R. Avissar g, A. Becker h, M. Claussen h, A.J. Dolman a, R.A. Feddes i, M. Fosberg j, Y. Fukushima k, J.H.C. Gash l, L. Guenni m, H. Hoff j, P.G. Jarvis m, I. Kayane n, A.N. Krenke o, Changming Liu p, M. Meybeck q, C.A. Nobre r, L. Oyebande s, A. Pitman t, R.A. Pielke Sr. u, M. Raupach v, B. Saugier w, E.D. Schulze x, P.J. Sellers y, J.D. Tenhunen z, R. Valentini aa, R.L. Victoria ab, C.J. Vörösmarty ac a DLO Winand Staring Centre for Integrated Land, Soil and Water Research, PO Box 125, NL 6700 AC, Wageningen, Netherlands b School of Forestry, University of Montana, Missoula, USA c Department of Hydrology and Water Resources, University of Arizona, Tuscon, USA d Department of Plant Biology, Carnegie Institution, Stanford, USA e Environment Canada, Downsview, Canada f Department of Astronomy and Geophysics, University of São Paolo, São Pãolo, Brazil g Rutgers University, New Brunswick, USA h Potsdam Institute for Climate Impact Research, Potsdam, Germany i Agricultural University, Department of Water Resources Management, Wageningen, Netherlands j BAHC Core Project Office, Potsdam Institute for Climate Impact Research, Potsdam, Germany k Institute for Hydrospheric Atmospheric Sciences, Nagoya University, Nagoya, Japan l Institute of Hydrology, Wallingford, UK m Universidad Simon Bolivar, Caracas, Venezuela n Institute of Geoscience, Tsukuba University, Tsukuba, Japan o Institute of Geography, Russian Academy of Sciences, Moscow, Russia p Institute of Agricultural Modernisation, Chinese Academy of Sciences, Shijiazhuang, People’s Republic of China q Université de Paris, Paris, France r INPE Center for Weather Prediction and Climate Studies, Cachoeira Paulista, Brazil s Hydrology Laboratory, University of Lagos, Lagos, Nigeria t Macquarie University, North Ryde, Australia u Colorado State University, Fort Collins, USA v Centre for Environmental Mechanics, CSIRO, Canberra, Australia w Lab d’Ecologie Vegetale, Université de Paris, Paris, France x Max Planck Institute for Biogeochemistry, Jena, Germany y NASA Goddard Spacer Flight Center, USA z Department of Plant Ecology, University of Bayreuth, Bayreuth, Germany a,a Universita della Tuscia, Viterbo, Italy a,b CENA, University of São Paolo, Piracicaba, Brazil a,c University of New Hampshire, Durham, USA * Corresponding author. Tel.: 131-317474744; 1 31-317424812; e-mail: hutjes@sc.dlo.nl Fax: 0022-1694/98/$ - see front matter q 1998 Elsevier Science B.V. All rights reserved. PII: S0022-169 4(98)00255-8 2 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Abstract The Core Project Biospheric Aspects of the Hydrological Cycle (BAHC) of the International Geosphere Biosphere Programme (IGBP) addresses the biospheric aspects of the hydrological cycle through experiments and modelling of energy, water, carbon dioxide and sediment fluxes in the soil-vegetation-atmosphere system at a variety of spatial and temporal scales. Active regulation of water, energy and carbon dioxide fluxes by the vegetation make it an important factor in regulating the Earth’s hydrological cycle and in the formation of the climate. Consequently, human induced conversion of vegetation cover is an important driver for climate change. A number of recent studies, discussed in this paper, emphasise the importance of the terrestrial biosphere for the climate system. Initially, these studies demonstrate the influence of the land surface on tropical weather and climate, revealing the mechanisms, acting at various scales, that connect increasing temperatures and decreasing rainfall to large-scale deforestation and other forms of land degradation. More recently, the significance of the land surface processes for water cycle and for weather and climate in temperate and boreal zones was demonstrated. In addition the terrestrial biosphere plays a significant role in the carbon dioxide fluxes and in global carbon balance. Recent work suggests that many ecosystems both in the tropics and in temperate zones may act as a substantial sink for carbon dioxide, though the temporal variability of this sink strength is yet unclear. Further, carbon dioxide uptake and evaporation by vegetation are intrinsically coupled, leading to links and feedbacks between land surface and climate that are hardly explored yet. Earth’s vegetation cover and its changes owing to human impact have a profound influence on a lateral redistribution of water and transported constituents, such as nutrients and sediments, and acts therefore as an important moderator of Earth’s biogeochemical cycles. In the BAHC science programme, the importance of studying the influence of climate and human activities on mobilisation and river-borne transport of constituents is explicitly articulated. The terrestrial water and associated material cycles are studied as highly dynamic in space and time, and reflect a complex interplay among climatic forcing, topography, land cover and vegetation dynamics. Despite a large progress in our understanding of how the terrestrial biosphere interacts with Earth’s and climate system and with the terrestrial part of its hydrological cycle, a number of basic issues still remain unresolved. Limited to the scope of BAHC, the paper briefly assesses the present status and identifies the most important outstanding issues, which require further research. Two, arguably most important outstanding issues are identified: a limited understanding of natural variability, especially with respect to seasonal to inter-annual cycles, and of a complex ecosystem behaviour resulting from multiple feedbacks and multiple coupled biogeochemical cycles within the overall climate system. This leads to two major challenges for the future science agenda related to global change research. First, there is a need for a strong multidisciplinary integration of research efforts in both modelling and experiments, the latter extending to inter-annual timescales. Second, the ever increasing complexity in characterisation and modelling of the climate system, which is mainly owing to incorporation of the biosphere’s and human feedbacks, may call for a new approach in global change impact studies. Methodologies need to be developed to identify risks to, and vulnerability of environmental systems, taking into account all important interactions between atmospheric, ecological and hydrological processes at relevant scales. With respect to the influence of climate and human activities on mobilisation and river-borne transport of constituents, the main issues for the future are related to declining availability and quality of ground data for quantity and quality of water discharge. Such assessments presented in this paper, in combination with community wide science evaluation, has lead to an update of the science agenda for BAHC, a summary of which is provided in the appendix. q 1998 Elsevier Science B.V. All rights reserved. Keywords: Hydrological cycle; Terrestrial biosphere; Global carbon balance; Natural variabilty; Biogeochemical cycle 1. Introduction About a decade ago the International Council of Scientific Unions established the International Geosphere Biosphere programme (IGBP, 1990). One of its core projects is Biospheric Aspects of the Hydrological Cycle (BAHC) (IGBP, 1993). BAHC was created to address the interactions between the vegetation and the physical processes of the hydrological cycle. The basic science agenda defined for BAHC involves: • Determination of the biospheric controls on the hydrological cycle through experiments and R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Fig. 1. Diagram showing the structure of IGBP core project Biospheric Aspects of the Hydrological Cycle, primarily based on scale. modelling of energy and water fluxes in the soilvegetation-atmosphere system at all spatial and temporal scales; • Development of appropriate data sets that describe the interactions between the terrestrial biosphere and the physical earth system, and that can be used for testing and validation of model simulations of these interactions. Through the active regulation of water and energy fluxes the terrestrial biosphere has an important influence on climate. A modification of the terrestrial biosphere, i.e. large-scale changes in land cover and land use, may lead to changes in climate. The relative importance of this particular driver of climate change is increasingly being recognised (Pielke et al., 1998; Kabat and Hutjes, 1998). Land cover change as a driver for climate change works in addition to, and interacts with climate change resulting from the alteration of the Earth’s radiation balance, caused by increased concentrations of greenhouse gases. Recent modelling studies suggest that major land cover changes in the past may have exacerbated, or even caused, the observed modifications of regional climate, which so far was ascribed to the greenhouse effect. Examples of such well-documented cases are the increased drought risks in the Mediterranean and the Sahel following removal of vegetation by forest clearing and over-exploitation respectively (Reale and Dirmeyer, 1998; Xue and Shukla, 1993). Current land cover changes, like tropical deforestation, are taking place at a much higher rate than ever occurred in the 3 past. These may not only influence climate directly, through the water and energy cycles, but as the process of conversion often includes biomass burning, they also add substantially to carbon dioxide emissions. Deforestation is the second major global source of atmospheric CO2 after emission of carbon by fossil fuel combustion (Houghton et al., 1996). Conversely, as growing vegetation acts also as a sink of carbon dioxide, land management in the form of afforestation has recently been proposed to become a new policy instrument to mitigate, i.e. temporise, further increases in carbon dioxide levels in the atmosphere (Steffen et al., 1998). The Earth’s vegetation cover and its modification owing to human impact have also a profound influence on a lateral redistribution of water and transported constituents, such as nutrients and sediments, and acts therefore as an important moderator of the Earth’s biogeochemical cycles. Large-scale land use changes, such as deforestation in the past, in combination with river regulation and with increasing levels of agriculture and industrial pollution, have produced a severe distortion of natural hydrographs and altered material transport (Vörösmarty et al., 1997a; Vörösmarty et al., 1997b; Humborg et al., 1997). Regional changes in the delivery of land-based constituents collectively appear to have imparted a biogeochemical signal at continental and global scales. For example, it was estimated that riverine transports of inorganic nitrogen and phosphorus have increased several fold over last 150–200 years (Meybeck and Helmer, 1989;Howarth et al., 1996). While these changes are accelerating, the basic monitoring of hydrological fluxes, through data provided by observational networks for discharge and water quality, remains insufficient and fragmented. It is therefore not surprising that we lack accurate quantitative estimate of continental to global-scale runoff, riverine constituent flux and the major factors that control these. Research on the impact of mankind on land cover and land use as an important driver of past, present and future climate change deserves a high priority. The BAHC programme seeks to foster and co-ordinate this particular research agenda and to identify the most progress-limiting issues as a basis for new, often multi-disciplinary research lines. This is an ongoing process, which also requires the programme itself to 4 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Fig. 2. Full suite of land surface (including ecological and biogeochemical) hydrology and atmosphere interactions active at various time scales (see text for further explanation). evolve. Initially, the BAHC programme was structured around four foci. These foci form a thematic structure based on a progression of time and spatial scales (see Fig. 1): 1. plot scale research on fundamental soil–vegetation–atmosphere interactions (Focus 1); 2. aspects of scaling and aggregation to regional scales as defined by the grid size of the mesoscale and global circulation models (GCM) (Focus 2); 3. long-term (seasonal to decadal) dynamics of coupling between biosphere, hydrology and climate (Focus 3); and 4. development of methods to disaggregate output of GCMs in a dynamically, physically and biologically consistent way to scales more appropriate for assessments of the impacts of global change on natural and man made ecosystems (Focus 4). The First Open Science conference of BAHC, convened as part of the EGS XX General Assembly in Hamburg, 1995, presented an overview of results, and new plans of a large number of research initiatives, contributing to several Foci’s of the BAHC research agenda. A selection of the papers presented at the conference was compiled in this volume and structured around the four BAHC research Foci’s. The conference provided an important contribution to a process of evaluation and refocusing of the BAHC science agenda. This process is near completion and it has resulted in a BAHC new science programme (Kabat, 1998) that seeks a higher level of integration. In the following section a number of recent studies that demonstrate the role and importance of the terrestrial biosphere in the functioning of the climate system are discussed. This is followed by a brief assessment of the present status and major deficiencies in our current understanding, as far as these fall within what BAHC considers its realm. R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Fig. 3. Schematic representation of various interactions between land surface and atmosphere. Three basic interaction mechanisms are depicted: (a) dotted lines, through the radiation balance; (b) broken lines, through energy partitioning and (c) continuous lines, through turbulence and wind convergence. Symbols: a , albedo; Qn, net radiation; H, sensible and l E, latent heat fluxes; z0, aerodynamic roughness length; u*, friction velocity; Ra, aerodynamic resistance; Rs, surface resistance; and Ts, surface temperature. See text for further explanation. This assessment is structured around the original four BAHC foci, and the conclusions are formulated in terms of an updated science implementation plan for BAHC, of which an outline is given in the appendix. At the same time, this assessment also provides a context for the papers included in present volume. An outline of a strategy to fill in the most pressing remaining gaps in our knowledge, as well as of a strategy to make optimal use of the existing knowledge base in support of policy formulations, concludes the paper. 5 several inter-linked mechanisms. Vegetation actively controls the evaporation through its internal physiology; which results in opening of the stomata in response to environmental conditions affecting photosynthesis, such as the temperature, humidity, radiation, CO2 concentration and soil moisture. The stomata respond rather quickly to a change in environmental conditions and a characteristic time scale for the coupling with the atmosphere is typically 0.01– 1 h (strong coupling). More passively, through its albedo and structure, vegetation also affects other terms of terrestrial energy balance, as well as the wind dynamics. Although the coupling between the surface characteristics, like the vegetation structure and the atmosphere, is a very fast turbulent process (typical time scales of seconds to hours), a significant change in albedo would typically follow the seasonal evolution of the vegetation (moderate coupling with time scales of a week to a year). Finally, the vegetation moderates the biogeochemical and hydrological cycles by changing the pools and the fluxes of nutrients, sediments and by its interference with a lateral redistribution of water and transported constituents. Vegetation and soils exchange large quantities of carbon dioxide and other greenhouse gases with the atmosphere, thereby influencing the overall radiation balance of the earth. The coupling of biogeochemical and soil chemical processes with the atmosphere is expected to be most significant at very long time scales (one to several hundreds years), but arguably, this coupling mechanism may be a most crucial component of the system, when investigating climatic and global change signals. 2. The coupling between the biosphere, atmosphere and the hydrological cycle 2.1. The impact of land surface on climate: energy and water links The processes within the terrestrial biosphere, atmosphere and the hydrological cycle are intrinsically coupled (see Fig. 2). The coupling mechanism in the biosphere–atmosphere system is non-linear and in all cases, bi- or multi-directional, which means that an individual component of the system is both under the influence of, as well as impacting upon, the remaining parts of the system. The strength of the coupling varies with the characteristic time scales, which range from seconds to millennia. The vegetation interacts with atmosphere by In Fig. 3, a schematic representation of various feedbacks between land surface and atmosphere is given for the three typical land surface characteristics: albedo, soil moisture and the aerodynamic roughness length, which is related to the height of the vegetation cover. In each loop, positive and negative feedbacks occur, which means that the overall effect on for example the rainfall will be a trade-off between these feedbacks. For instance, an increase in albedo, resulting from a land cover change, will lead to a 6 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Fig. 4. Sensitivity of global rainfall patterns to expansion of all the worlds deserts. Shown are rainfall anomalies of a ‘double desert’run as percentage of the control run. Results are from the last eight years out of a 10 year intergration with the Center for Ocean Land Atmosphere studies (COLA) GCM. Land surface is parameterised with the SsiB model. For the ‘double-desert’ case all ‘arid steppe’ (shrubs with bare soil) grid cells of the control were replaced by desert (bare soil) and some of the surrounding grid cells were replaced by ‘arid steppe’ (from Dirmeyer and Shukla, 1996) decrease of radiation absorption by the land surface, which in its turn will cause a decrease in latent and sensible heat fluxes to the atmosphere. This will further result in a decrease of cloudiness, less convergence and finally in lower rainfall. Lower rainfall means a decrease in surface soil moisture, which contributes to a further increase of the surface albedo (positive feedback). Lower cloudiness will also lead to an increase of incoming radiation, and owing to more available radiative energy at the surface, indirectly to an increase of latent and sensible heat fluxes (negative feedback). Over the past 20 years a number of studies have build a strong evidence on the effects of land R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 degradation on the Sahelian climate. Most of these studies were of the ‘force response’ type, where the response of climate to prescribed changes in land cover or to modifications of the land surface parameters, was analysed. In earlier studies, only the effect of changing a single land surface parameter was investigated, demonstrating that increases in albedo (Charney, 1975; Charney et al., 1977; Sud and Fennessey, 1982), decreases in soil moisture (Walker and Rowntree, 1977) and decreases in aerodynamic roughness (Sud et al., 1988), may lead to a substantial decline of rainfall in the area. Progress since then consisted mostly of increasingly realistic land surface parameterisations (for a review see Sellers et al., 1996b), in combination with more realistic land cover (change) scenarios. Unlike the earlier work these studies could well simulate the observed decadal rainfall anomalies in the Sahel (e.g. Xue and Shukla, 1993). Regional reductions of surface evaporation and moisture flux convergence, increased subsidence and shifts in the summer monsoon belt, were identified as the main large-scale causes of observed rainfall reductions. These large-scale studies revealed another interesting aspect of land surface climate forcing. In many studies teleconnections were observed, with climate changes not being limited to the area of land use change alone, but also manifesting itself well away in areas not (yet) disturbed. In a recent study Reale and Shukla (1998) show how Mediterranean deforestation in the Roman era led to a sharp reduction in rainfall over the Iberian Peninsula and Northwest Africa. However, also south of the Sahara, where land degradation occurred only two millennia later, experienced lower rainfall during the Roman period. At this scale the oceans play an important role, and the relative strength of forcing of land surface versus sea surface temperatures was subject to some studies. These suggest that in the Sahel, sea surface temperatures dominate at the seasonal to inter-annual time scale, and the land surface at the decadal time scale, (e.g.Rowell et al., 1995). At the mesoscale, patterns in land cover and soil moisture also become important (Mahfouf et al., 1988; see also Avissar, 1998). Thermal anomalies in a landscape (brought about by vegetation and/or soil moisture anomalies) may trigger thermal circulations, in turn triggering convection and leading to rainfall. In 7 semi-arid regions this mechanism may lead to preferred and seasonally persistent rainfall patterns (Taylor, 1998), as evaporation is highly dependent on strongly inhomogeneous antecedent rainfall. This is a form of land surface atmosphere feedback which occurs at a much smaller scale then previously thought, and therefore needs to be included in future regional studies of climate and global change. The general conclusions from these studies for the Sahel — reduction in rainfall after major vegetation clearing, both local and at larger scales — were confirmed for a number of other semi-arid to arid areas of the world (e.g.Xue, 1995), see Fig. 4 (from Dirmeyer and Shukla, 1996). Several studies were conducted on the potential impact of deforestation in the wet tropics. One of the first studies for the Amazon (Henderson-Sellers and Hornitz, 1984) predicted a 15–20 mm per month decrease in rainfall following a complete removal of tropical rainforest and its replacement by pasture. The GCM used in this study had no specific representation of the tropical forest, but the authors approximated rainforest vegetation by changing the albedo, the roughness length and the water holding capacity of a two-layer bucket soil model. Dickinson and Henderson-Sellers (1988) repeated the study with a more detailed description of the land surface. The improved land surface scheme performed better for evaporation than the older scheme, especially with respect to the diurnal cycle during both wet and dry season. However, evaporation was poor when compared against the measurements of Shuttleworth (1988). These and subsequent measurements (Gash et al., 1996) were used to improve the land surface description coupled to the Hadley GCM (Lean and Rowntree, 1996), which again predicted a roughly 30 mm/month decrease in rainfall following complete conversion of forest to pasture. However, recent mesoscale studies with realistic, partial forest conversion suggested that with limited, relatively dispersed deforestation rainfall reduction is much less severe and the rainfall might even increase, at least temporarily (IGBP, 1998). This is probably because of enhanced triggering of convection by contrasts in surface energy partitioning (forest vs. pasture), that occur in this particular area at a length scale similar to that at which convection is generated (see also Avissar and Liu, 1996; or Avissar, 1998). 8 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 There are two reasons that may explain why much of this work focuses on the tropics. First, at present land use changes occur mostly, and at unprecedented rates in the tropics, though in the past changes of similar magnitude occurred in other parts of the globe (Mediterranean, Central US, East and South Asia). The second reason is that at higher latitudes advection, i.e. frontal systems, dominate weather formation, and as such were a priori believed to mask any influence of the land surface. However, a number of more recent studies suggest that also under mid- and high-latitude conditions the land surface can play a strong role in weather formation. Improvements in prescribing seasonal vegetation dynamics (Dirmeyer, 1994), as well as using better land surface (soil) schemes (Pielke et al., 1996; Xue et al., 1996; Betts et al., 1996) improved weather prediction in the central US. Also, studies on anthropogenic land cover changes in the same region suggested marked influences on (near surface) climate (Copeland et al., 1996) and storm formation (Pielke et al., 1996). Tentative results by Van der Hurk (personal communication) indicate that also in Europe, at least in summer, land surface forcing of weather is significant. There are also indications of significant effects of land surface forcing at even higher latitudes. Betts and Ball (1997) showed how improvements in snow–albedo relationships, and in parameterising the influence of moss cover on soil evaporation enhanced the predictive capabilities of the ECMWF model. 2.2. Land surface–climate interactions: carbon and biogeochemical links Vegetation and soils play an important role as a store of carbon, while at the same time exchanging large quantities with the atmosphere. The world’s forests alone, representing about two-thirds of the total terrestrial carbon, contain about as much carbon as the atmosphere. The global rate of deforestation is 15–17 × 10 6 ha per year (Dixon et al., 1994;FAO, 1995) of which most is within the tropics and about half is in South America. Dixon et al. (1994) converted this areal loss into a mass of 0.9 Gt C per year, although some authors have estimated a much higher figure (Schimel, 1995). This makes deforestation the second major global source of atmospheric CO2, after emission of carbon by fossil fuel combustion, comparable to the sink strength of the total terrestrial vegetation (Houghton, 1996). The terrestrial biosphere exchanges large quantities of carbon with the atmosphere. Growing vegetation absorbs carbon dioxide for photosynthesis, the rate of which is strongly determined by available light, and soils release carbon dioxide through heterotrophic respiration, which is regulated mostly by temperature and soil wetness. These two fluxes are each more than an order of magnitude larger than the emission of carbon by fossil fuel combustion ( , 60 Gt/yr versus , 6 Gt/yr). However, the net balance of these two processes, the net terrestrial absorption of carbon, amounts to roughly 1–2 Gt/yr (Steffen et al.,1998). This net balance is difficult to assess, owing to a large temporal variability of both uptake and release processes in response to the variability of weather and climate, and as a result of large variety of ecosystems found on earth. Several recent studies have elucidated the processes that regulate uptake and release of CO2 at the annual to inter-annual time scale. Also, the changes in the global signal of seasonal CO2 concentrations (Keeling et al., 1996) show the sensitivity of climate variability and the consequences for source/sink relations of the biosphere (Walker et al., 1997). Studies using the NOAA flask network suggest that a significant portion of the uptake of the terrestrial biosphere is located at mid-latitudes (Denning et al., 1995). In this context, the tower flux data of EUROFLUX suggest that mid-latitude forests take up carbon at annual rates of 1–5 t/ha/yr (Valentini, 1998). However, also (primary) tropical forests may sequester carbon as recent, relative short-term Net Ecosystem Exchange (NEE) measurements in Amazonia indicate (Grace et al., 1995). Still, for the global carbon balance, the uncertainty involved in the stocks and fluxes of carbon in the tropics is significant (Schimel, 1995). A complicating factor in the biospheric control of climate is that the uptake of carbon dioxide by plants is strongly coupled to their regulation of evaporation. This coupling is based on the regulation of stomatal conductance of plants as a function of (amongst others) their photosynthesis rate (see also Cox et al., 1998). Many recent studies indicate that increased carbon dioxide concentrations, resulting from fossil fuel combustion, may ‘fertilise’ plants in the sense that they grow more efficient while using less water R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 9 Table 1 Comparison of radiative and physiological effects of doubled carbon dioxide on climate. Shown are changes in evapotranspiration, precipitation and surface air temperature averaged over the last ten years of 30 year GCM integrations, comparing current vs. doubled CO2 conditions. Three cases are compared: (R) simulation of the effects of CO2 increase on the radiation balance of the earth (the ‘classical’greenhouse effect); (PV) the effects of CO2 increase on plant photosynthesis and surface conductance, allowing for down-regulation of maximum photosynthesis rates in response to prolonged exposure of elevated CO2; and (RPV) simulation with the two effects combined. Shown are the effects in the tropics (14.48S–14.48N) where they are most pronounced. Adapted from (Sellers, 1996a) R PV RPV Evapotranspiration (%) Preciptation (%) Surface air temperature (8C) 5.1 2 4.2 2 0.8 5.0 2 0.2 2.1 0.7 1.7 2.6 (Walker et al., 1998). This fertilisation could feedback on climate in two opposing ways. On the one hand it was argued that this fertilisation may lead to a stronger sink strength, thereby mitigating atmospheric carbon dioxide increase and thus global warming. However, Schimel (1998) recently questioned the possibility for a larger uptake of CO2 by the biosphere under higher CO2 concentrations and a warmer climate, because that would also mean at least a doubling of nitrogen fluxes into ecosystems. Current nitrogen deposition trends do not indicate such increases, suggesting that future C-cycling will probably be more limited by nitrogen. On the other hand, as growth becomes limited by other factors than water, vegetation may sustain current growth rates at the expense of less water. Less evaporation implies smaller amounts of water are returned to the atmosphere, and at same time a larger fraction of the sun’s radiative energy is converted to warming the air. In a recent study, taking account of this effect in climate change prediction resulted in increased global warming rates comparable to ones owing to the conventional greenhouse effect, see Table 1 (from Sellers, 1996a). Though far from generally accepted, this result clearly suggests how the interplay between terrestrial vegetation, carbon dioxide emissions and the climate may alter our perception about the drivers of climate change. Therefore, it is of utmost importance to integrate the different biogeochemical cycles in a new generation of models, in which vegetation will be allowed to evolve, grow and die following seasonal and interannual cycle’s part of the climate, and interacting with the bio-geo-chemical cycles in the system. Yet, this new level of complexity may present us with major surprises. Modelling results show for example how the seasonality of vegetation affects the precipitation and temperature patterns at regional scales (Pielke, 1997), feeding back on LAI development. This suggests that feedbacks between terrestrial vegetation and the atmosphere amplify initial stochastic variability at the seasonal time scale. Also at much longer time scales feedbacks between vegetation and the climate system could amplify climatic changes. Some model studies suggests that the last glacial inception (e.g. de Noblet et al., 1996) and the warming during the mid-Holocene (e.g. Claussen and Gayler, 1997) are driven by changes in the Earth’s orbit, but they are amplified by vegetation–atmosphere interaction. Alternatively, vegetation-atmosphere models were found to converge to a single stable combination of climate and vegetation belts, irrespective of initial vegetation distribution in some eras, whereas in other eras differences in initial vegetation distribution lead to different stable states (Claussen, 1997a; Claussen, 1997b). 3. Biospheric controls of land surface–atmosphere interactions: progress and remaining issues 3.1. Plot scale processes From a global change perspective three general variants of Soil Vegetation Atmosphere Transfer (SVAT) models can be identified, all sharing a common set of core biophysical processes. The particular application of these variants determines the level of detail necessary to describe the core processes dealing with atmospheric and radiative transfer, soil moisture dynamics and its lateral redistribution, and canopy physiology and vegetation dynamics. SVATs 10 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 intended for incorporation in GCMs and numerical weather prediction models (NWPMs) basically focus on a correct representation of radiative energy exchange at the earth’ surface and its partitioning into latent and sensible atmospheric heat fluxes. Hydrology oriented SVATs are similar in their emphasis of energy and water exchange, but require a detailed understanding of soil moisture and runoff dynamics to predict correctly the hydrological routing and river discharges. Vegetation oriented SVATs put more emphasis on plant physiology and phenology to simulate ecosystem productivity and biogeochemical cycling. Besides these broad classes a wide range of SVATs were developed for more derived and/or geographically specific purposes. Arguably, the level of complexity of SVATs is determined by its intended applicability at a specific spatial and temporal scale. As the spatial scale increases a larger number of different ecosystems have to be described. This can be done either in a lumped fashion, employing a single parameterisation using effective, aggregated parameters, or in a distributed fashion using multiple parameter sets, or even seperate sub-models for each ecosystem (described later; also Avissar (1998) presents a discussion of such strategies). As the temporal scale increases the interactions between each of the basic components become more important to a point where all need to be explicitly modelled and coupled to each other. This aspect has made the current generation of land surface models to be more integrated descriptions of surface energy balance, photosynthesis, soil moisture and nutrient status. In particular, three areas currently attract most attention: ‘greening’ of SVATs through incorporation of more realistic plant physiology, ‘wetting’ of SVATs by improving on the wet hydrology parameterisations, and ‘generalisation’ by validation against long-term, inter-annual data sets from diverse biomes. Incorporation of carbon dynamics and vegetation physiology is motivated by a number of reasons. Arguably, better estimates of conductance can be achieved, with fewer empirical parameters, through the coupling of stomatal functioning and photosynthetic assimilation rates. From this coupling it also follows that vegetation is likely to respond directly to elevated CO2 levels by increases in productivity combined with a higher water-use efficiency, thereby enhancing the radiative climate response (Sellers, 1996a), as discussed in the previous section. In turn these changes may affect vegetation structure and functioning, and it is a matter of efficiency and consistency to use the same framework for modelling both the energy balance, and photosynthesis, and long-term plant structural and carbon dynamics. A number of papers in this volume fall in this area of research (Cienciala et al., 1998; Cox et al., 1998; Soegaard and Thorgeirsson, 1998). Studies in the context of the Project for Intercomparison of Land surface Parameterisation Schemes, PILPS, (Pitman and Henderson-Sellers, 1998) have shown a large variability in the ability of land surface parameterisations to simulate soil moisture. Analysing 14 different model codes currently in use in climate models, Wetzel et al. (1996) found that for all models runoff was systematically underestimated. They also found large differences between the various models. Further, incorporation of more physically based soil parameterisations led GCM and NWPM simulations to drift towards too dry climates. Thus, despite a great sensitivity of current climate and weather forecasts on soil moisture, the parameterisations of soil moisture and root water uptake are probably the weaker part of these models. As such, this remains an important area for future initiatives (covered by BAHC key theme two, see appendix). A number of papers in the present volume fall in this area (Yang et al., 1998; Callies et al., 1998). In the past, much developmental work on SVATs was based on observational data from one, or at best a few sites, spanning relatively short time periods. Large-scale applications of SVATs however, require validation and testing against long-term data sets, covering time scales from diurnal to multi-annual, from sites representative of the major biomes in the world. Observational techniques (e.g.Moncrieff et al., 1994) have now developed to the point where this need can be met by the establishment of long-term continuous flux measurement networks (Baldocchi et al., 1996). The so-called FLUXNET project (see BAHC key theme 1, appendix) co-ordinates the effort by encouraging standardisation, as far as appropriate, of observational techniques, sensors, calibration and auxiliary observations, as well as archiving of data in common accessible databases, and co-ordination of analysis efforts. Currently a large number of sites R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 are, or soon will be operational in Europe, the Americas, Australia and Asia. A number of recent SVAT intercomparison projects, like PILPS (Henderson-Sellers et al., 1996; Pitman and Henderson-Sellers, 1998), IGBP-GAIM’s Global Net Primary Productivity model intercomparison (Prince et al., 1995), and the Global Soil Wetness Index Project of GEWEX-ISLSCP (Dirmeyer et al., 1998), proved crucial in these developments. They identified both the most critical and/or most redundant processes and parameters. These projects developed a rational framework for data handling, model intercomparison and communication, that includes checks of internal consistency of the models against basic energy and mass closure, and perhaps more importantly, identifying the main areas of uncertainty in the current generation of SVAT models. A recent joint BAHC, ISLSCP, PILPS, GCTE workshop (Dolman and Dickinson, 1998) capitalised on these projects by addressing issues of parameterisation and data availability, with the general aim of setting out and prioritising lines for future development. Two issues were identified as crucial in virtually any SVAT component: the level of complexity required, and the necessity of treating the heterogeneity of the land surface (discussed later). Improvement of SVATs generally was accompanied by an increase in complexity. This raises problems in both the use of these models to represent large heterogeneous areas, as well as the methods and data required to test and calibrate these models. Thus, the need was identified for a more systematic study to define an optimal level of complexity of SVATs (BAHC key theme 3, see appendix, is a response to this need). 3.2. Landscape heterogeneity Real landscapes exhibit considerable heterogeneity at scales below the resolution of large-scale atmospheric models. This heterogeneity has aspects of topography, variation between and in vegetation types, and of variability in soil physical properties. It is important to understand when and how the presence of unresolved heterogeneity requires recognition in any aggregate representation. Aggregation here refers to a form of spatial averaging and involves two lines of research. When modelling spatially variable processes aggregation seeks to optimise the grid 11 size such that the resolution is fine enough to leave results unaffected by unresolved variability, yet coarse enough to allow efficient simulations. From a data perspective it searches for an adequate spatial resolution of measurements, requiring an assessment of the effects of resolution on the accuracy of remotely sensed data and derivatives. Land–atmosphere interactions and the issues of effective, aggregate representation of sub-grid variability were addressed in a series of large-scale field experiments (LSE) which were effectively conceived by the BAHC- and ISLSCP/GEWEX research communities. Examples of these are HAPEXMOBILHY (André et al., 1986), FIFE (Sellers et al., 1992) EFEDA (Bolle et al., 1993), HAPEX-Sahel (Goutorbe et al., 1994; Goutorbe et al., 1997), BOREAS (Sellers et al., 1995), NOPEX (Halldin et al., 1998) and many others. These projects each produced a wealth of data, and were the breeding ground for significant progress in many aspects of observational methodology and model development, not only from the large-scale perspective but also for plot-scale processes. The progress made in the field of aggregation was assessed during the Tuscon Aggregation Workshop (Michaud and Shuttleworth, 1997). It was concluded that (under particular circumstances) relatively simple aggregation techniques can be used in large-scale modelling over flat terrain. Simple aggregation rules for above ground surface properties were shown to provide surface fluxes within 10% of fluxes simulated by models resolving the full patch scale heterogeneity. However, dealing with the often very fine scale heterogeneity in soil hydraulic properties, soil moisture dynamics and root water uptake remains problematic (Kabat et al., 1997). In the past much effort was concentrated on vertical processes, and only recently the representation of lateral redistribution of soil moisture in large-scale models gains more attention (e.g. Becker, 1995). The effects of surface heterogeneity at the mesoscale is an area of continued research efforts. Surface induced mesoscale circulations require rather complex parameterisations that may involve a more direct coupling between SVAT and atmospheric boundary layer (ABL) models, a challenge taken up by BAHC key theme three (see appendix). The regional energy balance proved to be 12 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 insensitive to gentle topography, provided vegetation and soil moisture are uniform (Raupach and Finnigan, 1997). The representation of intensive topography and the strongly correlated variation in vegetation and soil (-moisture) in large-scale models requires an even more integrated approach to the problem than previous LSEs. BAHC key theme 7 was formulated in response to this need. Remotely sensed irradiances are relatively insensitive to unresolved variability. At the pixel level, both more directly derived variables (e.g. surface temperature), and simple indices that relate to plant physiological parameters (e.g. NDVI), seem to represent effective values representative of sub pixel variation (Moran et al., 1997). Mapping of more derived hydrometeorological variables through remote sensing is an area of very active research (examples in this volume: Prince et al., 1998; Bastiaansen et al., 1998; Schmugge, 1998). As many derived products are non-linearly dependent on primary observables (e.g. sensible heat flux on surface temperature), they are more sensitive to unresolved variability. A better mutual integration of SVAT models with remote sensing algorithms, and the combined use of multiple sensor types, short-wave, thermal and microwave, seems promising in this respect. 3.3. The largest scales Dealing with truly global environmental change issues requires extension of understanding, and thus of data and models, to larger spatial and temporal scales than those discussed in the two previous sections. At the same time this necessitates a new level of integration as more processes that can be treated more or less independently at small, especially temporal scales, interact strongly in the long run. Examples are biospheric controls on the surface energy balance through processes with increasingly longer ‘memories’ like soil moisture, nutrient availability and ambient carbon dioxide levels. In this context BAHC focussed in the past on three themes: global mapping of biome types, modelling of dynamic vegetation–climate interactions, and integration at catchment levels. Various initiatives aimed at producing global vegetation maps were implemented. The central problem is to map a continuous species distribution onto discrete classes, based on structural and/or functional criteria. Practical considerations (data availability) limit the number of classes that can be represented in land surface parameterisations for global models. In response to this limitation, attempts are being made to map continuous distributions of characteristics rather than discrete classes of biomes. An example of the effects of one approach versus the other is given by Martin (1998). Such continuous parameter maps instead of explicit classifications, are also more consistent with efforts to produce global aggregated land surface parameter maps (see new BAHC key theme 8, appendix) A recent BAHC-ISLSCP workshop on ‘Land Surface Data in Climate and Weather Models’, confirmed the importance of earth and its land masses for weather and climate, a role that is also increasingly recognised by major weather forecasting and climate prediction groups (Kabat and Hutjes, 1998). The review given in the previous sections of this paper are largely based on that workshop. The atmospheric modelling community increasingly takes notice of the impact of vegetation and its distribution and changes on climate. Contrarily, climate change impact studies implicitly focuses on the uni-directional effect of climate on natural and man made vegetations (as an example, see Kergoat, 1998). The BAHC-ISLSCP workshop on ‘Bi-directional Ecosystem Atmosphere Interactions’ (Field and Avissar, 1998) reviewed very recent attempts to couple climate and vegetation redistribution in a truly bi-directional way. Though the analyses are currently rather limited, the importance of bi-directional coupling was well recognised and its further development will be fostered by BAHC key theme 5 (see appendix). Impact studies of changes in climate and land cover on water resources, extreme events like floods and droughts, and land degradation through erosion and export of mineral and organic nutrients are all interlinked and may be studied through the behaviour of river drainage basins. The changing nature of river systems owing to human interventions is an important component of the global change debate. However, the development of large-scale river process models is impeded by too large uncertainties in current estimates of riverine fluxes of water and waterborne materials (e.g. Meybeck, 1994). R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 13 Fig. 5. Two alternative ways of downscaling GCM output: (a) dynamic downscaling in which the GCM provides the boundary conditions for higher resolution models; (b) stochastic downscaling, in which statistical relations between observed large-scale weather phenomena and near surface meteorology are applied to modelled large-scale patterns. Short-term progress could be made by a combination of several alternative modelling approaches, spatially explicit biophysical datasets, and improved river monitoring systems. A suitable river typology could permit extrapolation from data-rich catchments to larger domains. Also with respect to the internal dynamics of river systems new efforts are needed. These should be directed towards coupling of (largely existing) process-based models describing material erosion, transport, transformations and deposition along the entire continuum of river systems. The issue of feedbacks from changes of river systems on biogeochemical cycling, climate forcing, water resources need also be addressed in this integrated way. In both dynamics and feedback studies shortterm emphasis should be on case studies at the regional scale. BAHC key theme 6 will address such issues. 3.4. Downscaling Simulating ecohydrological impacts of climate change requires climate data at appropriate spatial and temporal scales. These data should, at least in a statistical sense, represent a series of individual weather events. The organisation of weather events is characterised by mesoscale atmospheric motions with typical sizes from a few to a hundred kilometres, and lifetimes ranging from hours to about a day. These scales are not resolved by GCMs, and not even by the observational network for large parts of the globe. GCMs may only reproduce, in a statistical sense, large-scale circulations in the order of a thousand kilometres and with lifetimes of multiple days. Developing methods to provide forcing data at high spatial and temporal resolution for climate change impact studies involves a number of research components. The first is the spatial-temporal analysis, interpolation and simulation of current, observed weather data. This provides part of the basis for the second component: the downscaling of GCM output. Both these issues were initially addressed mainly through statistical approaches. More recently, dynamical approaches are gaining increasing attention (see Fig. 5). Providing a physical basis for either stochastic and/or dynamic downscaling is central to a third component. Stochastic methods were developed and used to simulate weather variables at sites remote from monitoring stations, and/or for the analysis of extreme events under different weather types. This involves analysis of space–time covariance structures of associated weather anomalies. Spline techniques can be used to interpolate the parameters of such probability distributions at monthly time scales. Objective analysis seems more appropriate at the daily time scale. In more sophisticated approaches these probability distributions are conditioned on a classification of large-scale atmospheric flow (examples are Guenni and Hutchinson, 1998; Reiner and Lettenmaier, 1998; Wilby, 1998). The latter approach is also being used in stochastic downscaling of GCM output, where the weather pattern classification can be based on a suitable set 14 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 of predictor variables. Such predictor sets can include both free atmosphere variables (e.g. geopotential heights) and/or surface patterns (e.g. sea level pressure) (Karl et al., 1990). The basic assumption of this method, and one often criticised, is the invariance of the stochastic parameters under changed climates. This might be overcome with dynamical downscaling. Progress in modelling techniques and computing power over the last decade has enabled nesting of higher resolution models in particular areas of interest within GCMs. Coupling between the scales can be either one way, the GCM forcing the nested model, or bi-directional (Giorgi and Mearns, 1991; Russo and Zack, 1994). This physical coupling across scales ultimately provides the basis for both stochastic and dynamic approaches. Dynamics of the general circulation interact with synoptic scale processes, which in turn relate to convective scale processes that through boundary layer processes interact with the land surface. The bidirectional exchange of energy and mass between each of these scales is a current focal point of research (as example see Hantel and Acs, 1998). At each scale jump the variance in the low-resolution product can be reconstructed using the high-resolution physical boundary conditions from the next level. Downscaling plays an important role in translating global climate change scenarios to more regional risk assessments in support of policy formulations. However, a principally different approach might be needed as the much needed integration of more and more coupled, strongly interacting processes and biogeo-chemical feedbacks, may ultimately render climate prediction virtually impossible. Instead of impact research driven by climate change scenarios, it might be better to study susceptibility and vulnerability of different functions of the Earth system (water resources, ecosystem resilience, etc.) as part of a fully coupled earth system. Such a new strategy for ‘vulnerability’ assessment will be further developed under BAHC crosscutting theme B (see appendix). 4. Conclusions This paper briefly reviewed the role and importance of the land surface, i.e. terrestrial ecosystems, in climate formation. The land surface exerts its influence on climate through two major mechanisms: on the one hand through modification of the surface radiation balance, surface energy partitioning (evaporation) and wind fields and on the other hand through its exchange with the atmosphere of radiatively active (trace) gases, most notably carbon dioxide. Though the importance of this role gains increasing recognition by the climate and operational weather prediction communities, it receives little attention (at least until recently) in the policy debates around climate change. Illustrative of this neglect is absence of the issue in the ‘Summary for policy makers’ of the latest consensus report by the Intergovernmental Panel on Climate Change (Houghton et al., 1996). The relative importance of biospheric controls on climate increases because of the unprecedented rate of land cover conversions (deforestation, desertification, large-scale drainage or irrigation, etc.) of recent decades. It further increased because of the role land cover conversions (afforestation) have acquired in policies mitigating further carbon dioxide increases by fossil fuel combustion since the Kyoto negotiations on the Framework Convention on Climate Change. A number of recent sensitivity studies have shown the effects large-scale land cover conversion can have on weather and climate. Plant physiological control of evaporation affects surface energy partitioning and as a result boundary layer temperature, humidity and growth. In turn this can affect rainfall through its effects on the formation of clouds and on the triggering of convection. Ultimately it may also affect synoptic phenomena as the seasonal movements in polar and inter-tropical fronts. Thereby land surface anomalies exert their influence well beyond the region in which they occur. This influence occurs on practically all spatial and time scales, affecting both individual storm events as well as climate transitions following the last glaciation. It occurs at all latitudes, from tropics to sub-polar regions, and both in the continental interiors, as well as at their more marine influenced fringes. Contrarily terrestrial ecosystems exchange huge quantities of carbon dioxide with the atmosphere. Small perturbations in either uptake for photosynthesis, or release through heterotrophic respiration easily offset any scenarios for emissions from fossil R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 fuel combustion. Many mature ‘ecosystems’ at various latitudes were found to sequester carbon. However, given the limited time frame of such observations it is unclear whether their current sink strength is anomalous, resulting from short-term (inter-annual) climate variability, or more persistent in response to long-term environmental changes. The physiological response of vegetation to increased carbon dioxide levels is strongly coupled with nutrient availability. Given enough water and nutrients vegetation may grow faster, thus sequestering more carbon. Limited by nutrients it may grow while using less water thereby enhancing global warming owing to increased greenhouse gas levels. The role of adaptive strategies of ecosystems in response to a changing environment, migration or genetic adaptation has not been explored in any quantitative sense. Arguably, the two weakest issues in all these works are deficiencies in our understanding of natural variability at seasonal to inter-annual time scales, and of complex behaviour resulting from multiple feedbacks and multiple coupled biogeochemical cycles. The first is the result of the fact that most past experimental studies of land surface–atmosphere interactions were performed in campaign mode sampling a single season, or at best a few more in an intermittent fashion. The second results from the fact that most research in these fields was in the form of ‘force response‘ studies, observing or modelling the response of one system to changes in another. However, it becomes more and more widely recognised that the fundamental properties of the climate system, such as response mechanisms, regulation ability and adaptability, do not depend on the capacity of separate subsystems, or on a unidirectional driving of one component on the other. Rather, the earth system behaves as a dynamic entity, consisting of coupled, strongly interacting processes and biogeochemical cycles. These challenges have two major implications for the future science agenda related to global change. One is the need for a strong multi-disciplinary integration of research efforts in both modelling and experiments, extending the latter to longer time frames. The other is the need for a new approach in global change impact studies. To increase our understanding of the behaviour of complete ecosystems requires concurrent and colocated efforts from various disciplines. A good 15 example is the currently implemented Large-Scale Biosphere Experiment in Amazonia (LBA, Nobre et al., 1996). It differs from previous large-scale land surface experiments (Feddes et al., 1998) by its thematic integration, including components focussing on meteorology and hydrology, but also on carbon, biogeochemistry, atmospheric chemistry and on the socio-economic drivers of environmental change in that particular region. Also it takes a long-term perspective with many observations being made continuously for multiple years. BAHC will further promote this type of research as part of its crosscutting theme A. Increased understanding of complex systems does not guarantee we can also predict their fate. Climate change may become virtually unpredictable after inclusion in our models of all possible terrestrial and bio-geo-chemical feedbacks, with their additional non-linearities and thresholds. This would require an entirely new approach to climate change (impact) studies, focusing less on global means, trends and predictions, but more on regional rates, thresholds and resulting risks and options. Methodologies to assess risks and vulnerability need to be developed that take account of interactions and feedbacks between atmospheric, ecological and hydrological processes at different spatial and temporal scales. In the context of water resources this is the central issue in BAHC cross cutting theme B (see appendix). Having identified the ‘weak spots’ in the earth system in this way we will be better equipped to develop scenarios of future environmental change, and analyse them for threats. Appendix A. Future structure of IGBP core project Biospheric Aspects of the Hydrological Cycle To select from the huge societal demand for data, knowledge, scientific partnerships and policy advises, those issues that BAHC considers within its realm requires a filter. The same filter should serve as a tool to condense the output of a loosely defined research community, into focused products BAHC can justifiably claim to have played a role in. The following integrating question is central to the revised mission of BAHC: 16 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 How do changes in biospheric processes interact with global and regional climates, hydrological processes and water resources, when driven by changes in atmospheric composition and land cover? In order to achieve greater integration both within BAHC, and between BAHC and other co-ordinating programmes, its organisational structure will depart from loosening the original four foci division and put more emphasis on concrete tasks that often transcend scientific and programmatic boundaries. These considerations lead to the formulation of eight key themes that form the core of BAHC research. In addition, two integrative themes were formulated. Details of agendas, deliverables and responsible persons can be found in the new BAHC Science Implementation Plan to be published soon. The eight key themes are: 1. Energy, water and carbon fluxes at the patch scale — FLUXNET. This theme addresses the role of terrestrial ecosystems in regulating biosphere atmosphere exchanges of energy and mass. It focuses on long-term measurements of carbon dioxide, water and energy, and sometimes trace gas fluxes, in order to provide new insights on annual and inter-annual variability of biosphere– atmosphere exchange processes. Therefore, a longterm carbon and water flux monitoring network, FLUXNET, will be implemented and co-ordinated, with sites in the most relevant biomes across the globe (Baldocchi et al., 1996). 2. Evaluation of the role of below ground processes. Major anomalies in soil moisture affect the simulation of water and energy fluxes (Shukla and Mintz, 1982). However, the required accuracy of simulated soil moisture (in terms of quantity and spatial pattern) in climate models is unclear. Addressing this issue may help in prioritising the development of e.g. vertical vs. horizontal moisture transport. There is also a gap between the meaning of soil moisture in climate models and definitions of soil moisture that can be measured in the field. Related to this is the aggregation problem of soil hydraulic characteristics. Although still problematic, such issues were addressed for liquid soil moisture. However, the influence of frozen soil moisture received much less attention. In this contexts two issues will be addressed in this key theme: the parameterisation and monitoring of global-scale soil hydrology, and the parameterisation of root water uptake and the global mapping of root characteristics. 3. Parameterisation of land–atmosphere interactions. Despite large progress in SVAT modelling, there are still a number of basic problems in the parameterisation of the atmosphere exchange physics. Most SVAT models treat the surface as heterogeneous, accounting for bare soil, snow, vegetation and open water. Of the many components of SVAT models the modelling, especially at longer than seasonal time scales, of runoff, moisture availability (see previous theme) and vegetation physiology are arguably key issues. In addition, the heterogeneity as such is still an issue, as variability in vegetation and soil types interact with variability in precipitation to create large variations in energy fluxes. At the most basic level this key theme attempts to identify critical areas in SVAT modelling, evaluate performance and thus lead to improvement. 4. Land use climate interactions at the regional scale. Land cover and land use, and especially changes therein, interact at seasonal to multi-annual time scales and at more regional spatial scales. In this context a number of issues need to be explored. First, the strength of this direct coupling between land surface and climate at these relatively small scales need to be assessed for those regions where land use change is, or was strongest. Second, we have to understand how these regional, short-term interactions teleconnect to other regions and how they accumulate to more global modifications of climate. The role of slower, ‘memory-carrying’ aspects like soil moisture store, nutrient availability, etc. need more attention. The actual magnitude of stomatal control in response to weather and carbon dioxide enrichment needs better quantification, in combination with concurrent biogeochemical and ecosystem structural responses. 5. Global vegetation–climate interactions. The fundamental properties of the climate system, such as response mechanisms, regulation ability and adaptability, do not depend on the capacity of separate subsystems, or on a uni-directional driving of components on the other. Rather, the entire system behaves as a dynamic totality R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 consisting of coupled, strongly interacting processes. Therefore, BAHC will foster models that explicitly include the (bi-directional) interaction between global terrestrial vegetation and climate. Four issues need to be addressed. First, the strength of feedbacks within the entire climate system needs to be explored to assess the role of terrestrial vegetation in the climate system. Second, some vegetation–atmosphere models converge to a single stable combination of climate and vegetation belts, irrespective of initial vegetation distribution in some eras, whereas in other eras differences in initial vegetation distribution lead to different stable states (Claussen, 1997a; Claussen, 1997b). However, other models do not support this result and further analyses and model studies are needed to explore this problem. Third, a systematic investigation has to be undertaken to identify and explore the existence of hot spots in global vegetation–climate interactions. Fourth, the possibility that feedbacks between terrestrial vegetation and the climate system tend to amplify climatic changes owing to external forcing needs further analysis. 6. Influence of climate change and human activities on mobilisation and transport through riverine systems. During the last 18 000 years the architecture of river systems and the rates of fluvial water and material transports have changed dramatically. Causes are (a) climate and sea level variability, (b) human alteration of water fluxes and the mobilisation and transport of constituents from terrestrial ecosystems through land use conversion and channel modification. This key theme will analyse the key controlling factors that define the transport of water, particulate and dissolved material through river systems in the past, under contemporary conditions and into the future. Both current and selected past river basins should be mapped and characterised in a typological framework. This typology can then be used in conjunction with models and multi-regression techniques to identify key controlling factors. These models can be calibrated and validated using river discharge and water quality data sets from existing monitoring programmes such as WMO and GEMS/Water. 7. Mountain hydrology and ecology. This theme addresses the need to understand mountain 17 ecohydrological processes in the different mountain regions in the world. These processes are particularly diverse and sensitive to environmental change, and some related characteristics can be used as indicators of change, in connection with ecological and other indicators. Strategies for development should be derived which lead to sustainability and minimise adverse impacts of change. These issues are addressed in the recently introduced IGBP/IHDP Initiative on ‘Global Change Research in Mountain Regions‘ (IGBP, 1997). In this framework BAHC will focus on identification and long-term investigation of hydrological, including cryospheric, indicators of global environmental change, and on regional scale modelling framework for mesoscale studies along transects across different large mountain ranges and in associated complex mountain river basins. Also, attention will be given to small-scale studies of hydrological processes along altitudinal gradients. These should connect with studies in small mountain headwater basins, and to development of an integrated approach to analyse strategies for sustainable development in mountain regions taking, into account different scenarios of land use/cover and climate change. 8. Development of global data sets. Global, consistent parameter fields describing the land surface, i.e. vegetation and soils, are invaluable to those using large-scale meteorological, climatological, hydrological and/or ecological models. Several initiatives (WCRP-ISLSCP, IGBP-DIS) therefore aimed at producing such data sets for use in atmospheric or hydrologic models. The added value comes in the common reference these data sets provide against which to compare and test all kinds of (sub-) models. BAHC will play an active role in the conceptual design, production, and validation of global data sets of those parameters most relevant to its mission. Most of the key themes described before are either users of, or contributors to such data bases. More specifically, BAHC will contribute to the production of global aggregated land cover parameter data sets, and of global aggregated soil parameter data sets. Linked to the latter is the production of aggregated soil moisture time series for selected areas (centred on areas of completed LSEs). The design and production of a 18 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 global river inventory is a major component of key theme 6. The two crosscutting themes are: A. Design, prioritise and implement integrated terrestrial system experiments. A number of large-scale land surface experiments (LSE) were conducted over the past decade (Feddes et al., 1998). Most LSEs were conducted in campaign mode, covering at best a full growing season, and focussed on hydrometeorological (aggregation) issues in the land surface–atmosphere interactions. More recent experiments (BOREAS, Sellers et al., 1995) already extended their time frame as also more ecological aspects grew in importance. Currently in its implementation phase, the Large-Scale Biosphere Experiment in Amazonia (LBA, Nobre et al., 1996) is the example of the culmination of this trend to increase integration and extend observational time scales. BAHC has a clear role in the design, prioritisation and implementation of present and future ‘integrated terrestrial system experiments’ (ITSE). Therefore, BAHC will foster linkages and integration between co-located studies initiated from different, limited, disciplinary or programmatic perspectives. As initiatives are many, there is a clear need for prioritising ITSEs in consultation with both policy and scientific communities. Bringing together human, capital and funding resources to implement ITSEs is also a non-trivial task that requires dedicated attention in cases where BAHC relevant science forms a major part of such an ITSE. B. Scenario development and risk/vulnerability analysis. A change in global climate such as in the spatial and temporal distribution of precipitation and temperature, or significant changes in land use, will undoubtedly affect water. However, policy needs associated with global change are regional, are associated with risk, extremes, rates, thresholds and options. In contrast, scientists have emphasised global-scale changes, and resultant impacts in terms of means, trends and predictions (Bernabo, 1997). BAHC will address these issues by developing methodologies to assess risks and vulnerability with an explicit focus on fresh water resources (quality and quantity), that take account of interactions and feedbacks between atmospheric, ecological and hydrological processes at different spatial and temporal scales. Having identified the weak spots in the water system, BAHC will then develop scenarios of future environmental change, and analyse them for threats. Amongst others, this involves the development of high-resolution climate data. References André, J.C., Goutorbe, J.P., Perrier, A., 1986. HAPEX-MOBILHY: A hydrologic atmospheric experiment for the study of water budget and evaporation flux at the climatic scale. Bull. Am. Met. Soc. 67, 138–144. Avissar, R., 1998. Which type of SVATS is needed for GCMs. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4); 136–154. Avissar, R., Liu, Y., 1996. Three-dimensional numerical study of shallow convective clouds and precipitation induced by land surface forcing. J. Geophys. Res. 101, 7499–7518. Baldocchi, D.D., Valentini, R., Running, S., Oechel, W., Dahlman, R., 1996. Strategies for measuring and modelling carbon dioxide and water vapour fluxes over terrestrial ecosystems. Global Change Biology (Thematic Issue) 2.3, 159–168. Bastiaansen, W.G.M., Menenti, M., Feddes, R.A., Holtslag, A.A.M., 1998. A remote sensing surface energy balance algorithm for land (SEBAL): part 1, formulation; part 2, validation. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 198–212. Becker, A., 1995. Problems and progress in macroscale hydrological modelling. In: Feddes, R.A. (Ed.). Space and Time Scale Variability and Interdependencies in Hydrological Processes, Cambridge University Press, Cambridge, pp. 135. Bernabo, J.C., 1997. U.S. Climate Research and Assessments. Workshop Report Prediction in the earth sciences: use and misuse in policy making. Boulder, Colorado 10–12 July 1997. Betts, A.K., Ball, J.H., 1997. Albedo over Boreal forest. J. Geophys. Res. 102, 28 901–28 910. Betts, A.K., Ball, J.H., Beljaars, A.C.M., Miller, M.J., Viterbo, P., 1996. The land surface atmosphere interaction: a review based on observational and global modelling perpectives. J. Geophys. Res. 101, 7209–7225. Bolle, André, Arrue, Barht, Bessemoulin, Brasa, de Bruin, Cruces, Dugdale, , Engman, Evans, Fantechi, Fiedler, van de Griend, Imeson, Jochum, Kabat, Kratzsch, Lagouarde, Langer, Llamas, Lopez-Baeza, Wickland, Shuttleworth, Søgaard et al., 1993. EFEDA, European field experiment in a desertificationthreatened area, Ann. Geopysicae 11, 173–189. Callies, U., Rhodin, A., Eppel, D.P., 1998. A case study on variational soil humidity analysis from atmospheric observations. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1– 4), 95–108. R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Charney, J.G., 1975. Dynamics of deserts and drought in the Sahel. Quart. J. Roy. Meteorol. Soc. 101, 193–202. Charney, J.G., Quirk, W.K., Chow, S.H., Kornfield, J., 1977. A comparative study of the effects of albedo change on drought in semi arid regions. J. Atmos. Sci. 34, 1366–1385. Cienciala, E., Running, S.W., Lindroth, A., Grelle, A., Ryan, M.G., 1998. Analysis of carbon and water fluxes from the NOPEX boreal forest: Comparison of measurements with FORESTBGC simulations. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1– 4), 62–78. Claussen, M., 1997. Modeling biogeophysical feedback in the African and Indian Monsoon region. Climate Dyn. 13, 247–257. Claussen, M., 1997b. On multiple solutions of the atmosphere-vegetation system in present-day climate, Global Change Biol., in press. Claussen, M., Gayler, V., 1997. The greening of Sahara during the mid-Holocene: results of an interactive atmosphere – biome model. Global Ecology and Biogeography Letters, in press. Copeland, J.H., Pielke, R.A., Kittel, T.G.F., 1996. Potential climatic impacts of vegetations change: A regional modeling study. J. Geophys. Res. 101, 7409–7418. Cox, P.M., Huntingford, C., Harding, R.J., 1998. A canopy conductance and photosynthesis model for use in GCM land surface scheme. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1– 4), 79–94. Denning, A.S., Fung, I.Y., Randall, D., 1995. Latitudinal gradient of atmospheric CO2 due to seasonal exchange with land biota. Nature, vol. 376, 240–243. Dickinson, R.E., Henderson-Sellers, A., 1988. Modelling tropical deforestation: a study of GCM land surface parameterisations. Quart. J. Roy. Meteorol. Soc. 114 /volume-nr . , 58–72. Dirmeyer, P.A., 1994. Vegetation stress as a feedback mechanism in mid latitude drought. J. Climate 7, 1463–1483. Dirmeyer, P. A., Dolman, A.J., Sato, N., 1998. The global soil wetness project: A pilot project for global land surface modeling and validation. Bull. Amer. Meteor. Soc. in preparation. Dirmeyer, P., Shukla, J., 1996. The effect on regional and global climate of expansion of the world’s deserst. Quart. J. Roy. Meteorol. Soc. 122, 451–482. Dixon, R.K., Brown, S., Houghton, R.A., Solomon, A.M., Trexler, M.C., Wisniewski, J., 1994. Carbon pools and flux of global ecosystems. Science 263, 185–190. Dolman, A.J., Dickinson, R.E., 1998. Soil vegetation atmosphere models and land surface parameterizations: conclusions and working group reports of an IGBP/GEWEX workshop at La Jolla, CA, February 1997 khttp://www.cais.com/gewex/lspsvatrept.htmll. FAO, 1995. Forest Resources Assessment 1990, Global Synthesis, FAO Forestry Paper 124, FAO, Rome, Italy. Feddes, R.A., Kabat, P., Dolman, A.J., Hutjes, R.W.A., Waterloo, M.J., 1998. Large scale field experiments to improve land surface parameterisations. In Dooge, J.H. et al. (Eds.), Climate and Water: A 1998 perspective, in press. Field, C.B., Avissar, R., 1998. Bidirectional interactions between biosphere and the atmosphere. Global Change Biology, in press. 19 Gash, J.H.C., Nobre, C.A., Roberts, J.M., Victoria, R.L. (Eds.), 1996. Amazonian Deforestation and Climate Wiley, Chichester. Giorgi, F., Mearns, L.O., 1991. Approaches to the simulation of climate change: a review. Rev. Geophys. 29, 493–501. Goutorbe, Lebel, Tinga, Bessemoulin, Brouwer, Dolman, , Engman, Gash, Hoepffner, Kabat, Kerr, Monteny, Prince, Said, Sellers, Wallace, 1994. HAPEX-Sahel: a large-scale study of land– atmosphere interactions in the semi-arid tropics. Annales Geophysicae 12 53–64. Goutorbe, J.P., Dolman, A.J., Gash, J.H.C., Kerr, Y.H., Lebel, T., Prince, S.D., Stricker, J.N.M. (Eds.), 1997. HAPEX-Sahel. Special Issue of The Journal of Hydrology, Vols. 188–189. Grace, J., Lloyd, J., McIntyre, J., Miranda, A.C., Meir, P., Miranda, H.S., Moncrieff, J.M., Massheder, J., Nobre, C., Wright, I.R., Gash, J., 1992. Carbon dioxide uptake by an undisturbed tropical rain forest in South-West Amazonia 1992–1993. Science 270, 778–780. Guenni, L., Hutchinson, M.F., 1998. Spatial interpolation of the parameters of a rainfall model from groundbased data. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 335–347. Halldin, S., Gottschalk, L., van der Griend, A., Gryning, S.E. et al., 1998. NOPEX – A northern hemisphere climate processes land–surface experiment. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, this volume. Hantel, M., Acs, F., 1998. Physical aspects of the weather generator. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 393–411. Henderson-Sellers, A., McGuffie, K., Pitman, A.J., 1996. The project for intercomparision of land surface parameterization schemes (PILPS): 1992–1995. Climate Dynamics 12, 849–859. Henderson-Sellers, A., Hornitz, A., 1984. Possible climatic impacts of land cover transformations with particular emphasis on tropical deforestation. Clim. Change 6, 231–258. Houghton, J.T. (Ed.), 1996. Climate Change, 1995: The Science of Climate Change Cambridge University Press, Cambridge. Howarth, R., et al., 1996. Regional nitrogen budgets and riverine N and P fluxes for the drainages to the North Atlantic Ocean: Natural and human influences. Biogeochemistry 35, 75–139. Humborg, C., Ittekkot, V., Coclasu, A., Bodungen, B.v., 1997. Effect of Danube River dam on Black Sea biogeochemistry and ecosystem structure. Nature 386, 385–388. The International Geosphere Biosphere Programme: a study of global change, The International Geosphere Biosphere Programme (IGBP). The initial core projects, Report No. 12, IGBP, 1990, Stockholm. Biospheric Aspects of the Hydrological Cycle. The operational plan, Report No. 27, The International Geosphere Biosphere Programme (IGBP), 1993, Stockholm. Predicting global change impacts on mountain hydrology and ecology: integarted catchment hydrology and altitudinal gradient studies. Report No. 43, The International Geosphere Biosphere Programme (IGBP), 1997, Stockholm. LBA is moving forward. Global Change Newsletter 33, The International Geosphere Biosphere Programme (IGBP), 1998, Stockholm, 1–9. 20 R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Kabat, P., Hutjes, R.W.A., 1998. Mankind’s impact on Earth’s land cover as important for the climate as the oceans and industrial emissions of greenhouse gases khttp://www.cais.com/gewex/ pressreleaseapril98.htmll. Kabat, P. et al., 1998. Biospheric Aspects of the Hydrological Cycle: science and implementation plan 1998–2003. IGBP, Stockholm, in press. Kabat, P., Hutjes, R.W.A., Feddes, R.A., 1997. The scaling characteristics of soil parameters: from plot scale heterogeneity to subgrid parameterization. J. Hydrology 190, 363–396. Karl, T.R., Wang, W.C., Schlesinger, M.E., Knight, M.E., Portman, D., 1990. A method of relating general cicrulatioin model simulated climate to the observed local climate, part I: seasonal statistics. J. Climate 3, 1053–1079. Keeling, C.D., Chin, J.F.S., Whorf, T.P., 1996. Increased activity of northern vegetation inferred from atmospheric CO2 measurements. Nature 382, 146–149. Kergoat, L., 1998. A model for hydrologic equilibrium of leaf area index at the global scale. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 268–286. Lean, J., Rowntree, P.R., 1996. Understanding the sensitivity of a GCM simulation of Amazonian deforestation to the specification of vegetation and soil characteristics. Climate Research Technical Note 65, Hadley Centre, UK. Mahfouf, J.-F., Richard, E. et al., 1987. The influence of soil and vegetation on the development of mesoscale circulations. Journal of Climate and Applied Meteorology, 26, 1483–1495. Martin, P.H., 1998. Land-surface characterization in climate models: biome-based parameter inference is not equivalent to local direct estimation. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 287–303. Meybeck, M., 1994. Origin and variable composition of present day riverborne material. In: Material and Fluxes on the Surface of the Earth. Nat. Res. Council, Washington DC, pp. 61–73. Meybeck, M., Helmer, R., 1989. The quality of rivers: From pristine stage to global pollution. Paleogeography, Paleoclimatology, Paleoecology (Global and Planetary Change Section) 75, 283–309. Michaud, J.D., Shuttleworth, W.J., 1997. Aggregate description of land atmosphere interactions. J. Hydrology 190, 176–181. Moncrieff, J.B., Massheder, J.M., De Bruin, H.A.R., Elbers, J., Friborg, T., Heusinkveld, B., Kabat, P., Scott, S., Soegaard, H., Verhoef, A., 1994. A system to measure surface fluxes of momentum, sensible heat, water vapour and carbon dioxide. J. Hydrology 188-189, 589–611. Moran, M.S., Humes, K.S., Pinter Jr., P.J., 1997. The scaling characteristics of remotely sensed variables for sparsely vegetated heterogeneous landscapes. J. Hydrology 190, 337–362. de Noblet, N., Prentice, I.C., Jousaumme, S., Texier, D., Botta, A., Haxeltine, A., 1996. Possible role of atmosphere-biosphere interactions in triggering the last glaciation. Geophys. Res. Lett. 23, 3191–3194. Nobre, C.A., Dolman, A.J., Gash, J.H.C., Hutjes, R.W.A., Jacob, D.J., Janetos, A.C., Kabat, P., Keller, M., Marengo, J.A., McNeal, R.J., Melillo, J., Sellers, P.J., Wickland, .E., Wofsy, S.C., 1996. The large scale biosphere atmosphere experiment in Amazonia, concise experimental plan. LBA Project Office, Wageningen, Netherlands. Pielke, R.A., 1997. Role of land cover as a driving force for regional climate change. In: van de Geijn, S.C. Kuikman, P.J. (Eds.), International Workshop on Prospects for Coordinated Activities in IGBP Core Projects GCTE, BAHC and LUCC. Wageningen, Netherlands. Pielke, R.A., Lee, T.J., Copeland, J.H., Eastman, J.L, Ziegler, C.L, Finley, C.A., 1996. Use of USGS provided data to improve weather and climate simulation. Ecol. Applic. 7, 3–21. Pielke, R.A., Avissar, R., Raupach, M.R., Dolman, A.J., Zeng, X., Denning, S., 1998. Influence of short and long term ecosystem dynamics on weather and climate. Global Change Biology 4, 461–476. Pitman, A.J., Henderson-Sellers, A., 1998. Recent progress and results from the project for intercomparison of land surface parameterization schemes. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 128–135. Prince, S.D., Goetz, S.J., Dubayah, R.O., Czajkowski, K.P., Thawley, M., 1998. Inference of surface and air temperature, atmospheric precitiable water and water vapour deficit using AVHRR satellite observations: validation of algorithms. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 230–250. Prince, S.D., Olson, R.J., Dedieu, G., Esser, G., Cramer, W., 1995. Global Primary Production Data Initiative Project Description. IGBP-DIS Working Paper No. 12, IGBP-DIS, Toulouse, France. Raupach, M.R., Finnigan, J.J., 1997. The influence of topography on meteorological variables and surface–atmosphere interactions. J. Hydrology 190, 182–213. Reale, O., Dirmeyer, P, 1998. Modelling the effects of vegetation on mediterranean climate during the Roman Classical Period. Part I. Climate history and model sensitivity, J. Climate, submitted. Reale, O., Shukla, J., 1998. Modelling the effects of vegetation on mediterranean climate during the Roman Classical Period. Part II: high resolution model simulation, J. Climate, submitted. Rowell, J.P., Folland, C.K., Maskell, C., Ward, M.N., 1995. Variability of tropical rainfall over tropical North Africa (1906–92): observation and modelling. A. J. R. Met. Soc. 121, 669–704. Russo, J.M., Zack, J.W., 1994. Downscaling GCM output with a mesoscale model. In: Paoli, G. (Ed.), Climate Change, Uncertainty and Decision Making, pp. 47–60. Schimel, D.S., 1995. Terrestrial ecosystems and the global carbon cycle. Global Change Biology 1, 77–91. Schimel, D.S., 1998. Climate change: the carbon equation. Nature 393, 208–209. Schmugge, T., 1998. Applications of passive microwave observations of surface soil moisture. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 188–197. Schnur, R., Lettenmaier, D.P., 1998. A case study of statistical downscaling in Australia using weather classification by recursive partitioning. Special Issue of The Journal of R.W.A. Hutjes et al. / Journal of Hydrology 212–213 (1998) 1–21 Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 362–379. Sellers, Bounoua, Collatz, Randall, Dazlich, Los, Berry, Fung, Tucker, Field, Jensen,1996a. Comparison of radiative and physiological effects of doubled atmospheric CO2 on climate. Science 271,1402–1406. Sellers, P.J., Dickinson, R.E., Randall, D.A., Betts, A.K., Hall, F.G., Berry, J.A., Collatz, G.J., Denning, A.S., Mooney, H.A., Nobre, C.A., Sato, N., Field, C.B., Henderson-Sellers, A., 1996b. Modelling the exchanges of energy, water and carbon between continents and the atmosphere. Science 275, 502–509. Sellers, P.J., Hall, F.G., Asrar, G., Strebel, D.E., Murphy, R.E., 1992. An overview of the first international satellite land surface climatology project (ISLSCP) field experiment FIFE. J. Geophys. Res. 97 (d17), 18 345–18 372. Sellers, P.J., Hall, F., Margolis, H., Kelly, B., Baldocchi, D., den Hartog, G., Cihlar, J., Ryan, M.G., Goodison, B., Crill, P., Ranson, K.J., Lettenmaier, D., Wickland, D.E., 1995. The Boreal ecosystem-atmosphere study (BOREAS): An overview and early results from the 1994 field year. Bull. Am. Met. Soc. 76 (9), 1549–1577. Shukla, J., Mintz, A., 1982. The influence of land surface evaporation on the earth’s climate. Science 247, 1322–1325. Shuttleworth, W.J., 1988. Evaporation from Amazonian rainforest. Proc. Roy. Soc. 233, 321–346. Soegaard, H., Thorgeirsson, H., 1998. Carbon dioxide exchange at leaf and canopy scale above agricultural crops in the boreal environment. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1– 4), 51–61. Steffen, W., Noble, I., Canadell, J., Apps, M., Schulze, E.D., Jarvis, P.G., Baldocchi, D., Ciais, P., Cremer, W., et al., 1998. The terrestrial carbon cycle: implications for the Kyoto protocol. Science 280, 1393–1394. Sud, Y.C., Fennessey, M., 1982. A study of the influence of surface albedo on July circulation in semi arid regions using the GLAS GCM. J. Climatology 2, 105–125. Sud, Y.C., Shukla, J., Mintz, Y., 1988. Influence of land surface roughness on atmospheric circulation and rainfall: a sensitivity study with a general circulation model. J. Appl. Met. 27, 1036– 1054. Taylor, C.M., Lebel, T., 1998. Observational evidence of persistent convective-scale rainfall patterns, Mon. Wea. Rev., vol. 126, no. 6, 1597–1607. Valentini, R., 1998. Euroflux 2nd Year Annual Report. Vörösmarty, C.J., Sharma, K., Fekete, B., Copeland, A.H., Holden, 21 J., Marble, J., Lough, .A., 1997. The storage and aging of continental runoff in large reservoir systems of the world. Ambio 26, 210–219. Vörösmarty, C.J., Meybeck, M., Fekete, B., Sharma, K., 1997. The potential impact of neo-Castorization on sediment transport by the global network of rivers. In: Walling, D., Probst, J.-L. (Eds.). Human Impact on Erosion and Sedimentation, IAHS Press, Wallingford, UK, pp. 261. Walker, B., Steffen, W., Bondeau, A., Bugmann, H., Campbell, B., Canadell, P., Author, C., Cramer, W., Ehleringeer, J., Elliot, T., Foley, J., Gardner, B., Goudriaan, J., Gregory, P., Hall, D., Author, H., Ingram, J., Körner, C., Landsberg, J., Langridge, J., Lauenroth, B., Leemans, R., Linder, S., McMurtrie, R., Menaut, J.C., Mooney, H., Murdiyarso, D., Noble, I., Parton, B., Pitelka, L., Ramakrishnan, K., Sala, O., Scholes, B., Schulze, D., Shugart, H., Stafford-Smith, M., Suthurst, B., Valintin, C., Woodward, I., Zhang, X.S., 1998. The terrestrial bioshere and global change. Implications for natural and managed ecosystems. A synthesis of GCTE and related research. November 1997, IGBP Science Stockholm, p. 32. Walker, J., Rowntree, P.R., 1977. The effects of soil moisture on circulation and rainfall in a tropical model. Quart. J. Roy. Meteorol. Soc. 103, 29–46. Wetzel, P.J., Lian, X., Itranjad, P., Bone, A., Noilhan, J., Shao, Y., Skelly, C., Xue, Y., Yang, Z.L., 1996. Modelling vadose zone liquid water fluxes: infiltration, runoff, drainage, interflow. Glob. Plan. Change 13, 57–72. Wilby, R.L., 1998. Modelling low frequency rainfall events using airflow indices, weather patterns and frontal frequencies. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 380–392. Xue, Y., 1995. The impact of desertification in the Mongolian and the inner Mongolian grassland on the regional climate. J. Climate. Xue, Y., Fennessy, M.J., Sellers, P.J., 1996. Impact of vegetation properties on US weather prediction. J. Geophys. Res. 101, 7419–7430. Xue, Y., Shukla, J., 1993. The influence of land surface properties on Sahel climate, part I: desertification. J. Climate 4, 345–364. Yang, Z.-L., Dickinson, R.E., Shuttleworth, W.J., 1998. Treatment of soil, vegetation and snow in land-surface models: A test of the biosphere-atmosphere transfer scheme with the HAPEXMOBILHY, ABRACOS and Russian data. Special Issue of The Journal of Hydrology on Biospheric Aspects of the Hydrological Cycle, 212–213 (1–4), 109–127.