Physics and Chemistry of the Earth 30 (2005) 334–338 www.elsevier.com/locate/pce
Contemporary ‘‘green’’ water ﬂows: Simulations with a dynamic global vegetation and water balance model Dieter Gerten
, Holger Hoﬀ b, Alberte Bondeau a, Wolfgang Lucht a, Pascalle Smith c, So¨nke Zaehle a
a Potsdam-Institute for Climate Impact Research, Telegrafenberg, D14412 Potsdam, Germany University of Potsdam, Department of Plant Ecology and Nature Conservation, Maulbeerallee 2, D–14469 Potsdam, Germany c Laboratoire des Sciences du Climat et de l’Environnement LSCE, Bat. 709 Orme des merisiers CEA Saclay, F–91191 Gif-sur-Yvette, France
Available online 20 July 2005
Abstract ‘‘Green water’’—the water stored in the soil and productively used for plant transpiration—is an important quantity particularly in rainfed agriculture (in contrast to ‘‘blue water’’ available in streams and lakes, on which irrigation relies). This study provides preliminary estimates of contemporary (1961–1990) global green water ﬂows (i.e. plant transpiration), using a well-established, process-based dynamic global vegetation and water balance model, LPJ. Transpiration is analysed with respect to diﬀerences between a simulation that accounts for human land cover changes and a simulation under conditions of potential natural vegetation. We found that historic land cover change usually reduced the green water ﬂow to the atmosphere, resulting in a global decrease of 7% in total. To further explore how the biophysical setting inﬂuences the green water ﬂow, we analyse the ratio between soil moisture-limited canopy conductance of carbon and water and energy-controlled potential conductance. This plant physiology-based ratio measures the degree to which actual green water ﬂow falls below the potential ﬂow that would occur when the soil is saturated, thus it represents a measure of the water limitation of terrestrial vegetation. We found that plant water limitation is lowest in the wet tropics and at high latitudes, where soil moisture is high enough to meet the atmospheric demand for transpiration. The present results are preliminary, since irrigation is not yet accounted for, and because the model simulations are compromised primarily by the quality of the input data. A more comprehensive and consistent assessment of global green and blue water ﬂows and limitations using an enhanced LPJ model is identiﬁed as a prime task for future studies. Ó 2005 Elsevier Ltd. All rights reserved. Keywords: Global water resources; Soil moisture; Transpiration; Ecohydrology; Water stress; LPJ model
1. Introduction Humans are perturbing the terrestrial water cycle at an unprecedented rate, both directly through diversions and withdrawals and indirectly e.g. through land cover conversions and anthropogenic climate change (LÕvovich and White, 1990). At the same time, the number
Corresponding author. Fax: +49 3 31 2 88 2640. E-mail address: [email protected]
1474-7065/$ - see front matter Ó 2005 Elsevier Ltd. All rights reserved. doi:10.1016/j.pce.2005.06.002
of people aﬀected by water stress is increasing, a trend that will likely continue into the future (e.g. Cosgrove and Rijsberman, 2000). In order to understand the effects of human activities, it is imperative to investigate the potential pathways of water and their temporal changes from diﬀerent perspectives, and at global scale. For classifying the origin of water used for human purposes, one can distinguish between ‘‘blue water’’ and ‘‘green water’’, a concept that is important especially in food production (Falkenmark, 1997). Blue water refers to the water in rivers, lakes and aquifers that can potentially be withdrawn e.g. for industrial use and
D. Gerten et al. / Physics and Chemistry of the Earth 30 (2005) 334–338
particularly for irrigation. Green water is the precipitation water stored in the soil and eventually transpired by natural and agricultural vegetation, i.e. used for plant growth and biomass production. Unproductive evaporation e.g. from bare soil, vegetation canopies, or open waterbodies can be referred to as ‘‘white water’’ (Savenije, 2002). About 60% of all food globally is produced under non-irrigated rainfed conditions, that is, with green water (Cosgrove and Rijsberman, 2000). As green water does (and likely will continue to) play a key role for increasing food production in the future, it is important to assess, quantitatively, its spatial and temporal patterns. Here we quantify global green water ﬂows for both potential natural vegetation (biomes) and for cropland over the period 1961–1990. The analysis is based on simulations from the Lund–Potsdam–Jena (LPJ) dynamic global vegetation and water balance model (Sitch et al., 2003; Gerten et al., 2004a). The results extend previous estimates of green water ﬂows (Rockstro¨m and Gordon, 2001; Gordon et al., 2003a,b) in that these ﬂows are computed in a process-based manner considering the inﬂuences of e.g. climatic variations and CO2 concentrations upon them. In the context of ‘‘balancing water for humans and nature’’ (Falkenmark and Rockstro¨m, 2004), it is not only important to know how much green water is released into the atmosphere but also how strongly natural vegetation and crops are water-limited. As a contribution to this discussion, this study also investigates the degree to which transpiration (green water ﬂow) and photosynthesis of natural vegetation are co-limited by soil moisture deﬁcits.
2. The Lund–Potsdam–Jena model (LPJ) LPJ is a coupled biogeography–biogeochemistry and hydrology model of intermediate complexity, with process-based representations of terrestrial vegetation dynamics and land–atmosphere carbon and water exchanges at a spatial resolution of 0.5°. The model has been designed for simulating transient changes in major vegetation types in response to variations in e.g. climate and atmospheric CO2 content (McGuire et al., 2001; Lucht et al., 2002; Sitch et al., 2003), including feedbacks between vegetation and the water cycle. LPJ performs at the level of state-of-the-art global hydrological models with respect to the quality of simulated runoﬀ, and it reproduces well observed ﬁelds of soil moisture and evapotranspiration (Wagner et al., 2003; Gerten et al., 2004a). Key ecosystem processes considered in the model are seasonal vegetation growth, primary production, plant water productivity, mortality, carbon allocation, and resource competition. To account for variations in structure and function among natural and agricultural
plants, 10 plant functional types (Sitch et al., 2003), 11 crop types and two pasture types are distinguished (Bondeau et al., unpublished manuscript). Biomass production is computed based on a coupled photosynthesis–water balance scheme that explicitly considers the mutual dependence of transpiration and carbon uptake. The presence and fractional cover of natural plant types is determined annually for each grid cell according to individual bioclimatic, physiological, morphological, and ﬂammability attributes. The geographical distribution of crop types and pastures is derived from a global land use data set (IMAGE 2.2, IMAGE-team, 2001), and the annual fractional cover of a grid cell is taken from a historical croplands dataset (Ramankutty and Foley, 1999). Seasonal phenology of natural vegetation is determined according to the progression of temperature and soil moisture (Sitch et al., 2003). Crop phenology is determined from a temperature-dependent sowing date and growing degree-day requirements, also accounting for water-stress impacts. Soil moisture is calculated as the balance between the amount of water inﬁltrating into the soil (snowmelt, and precipitation minus interception loss from canopies) and that removed from two deﬁned soil layers (upper, 50 cm; lower, 100 cm thick) through surface and subsurface runoﬀ, percolation, and evapotranspiration. Relative soil moisture (Wr) is expressed as the fraction of current soil water content compared to texture-dependent plantavailable ﬁeld capacity. Wr thus measures the soil water deﬁcit, and it ranges between 0 and 1. Transpiration is calculated as the lesser of soil water supply (determined by potential evapotranspiration according to the Priestley–Taylor method, Wr, and rooting depth) and atmospheric demand. The demand is essentially a function of potential evapotranspiration and potential canopy conductance, which in turn is controlled by photosynthesis rate and ambient CO2 concentration. If soil water supply is lower than demand, canopy conductance, transpiration, and photosynthesis are reduced simultaneously, following Wr (Sitch et al., 2003; Gerten et al., 2004a). Thus, we use the ratio between actual and potential canopy conductance, denoted L, as a measure of the degree to which both photosynthesis and transpiration are constrained when a soil moisture deﬁcit exists (i.e. when Wr < 1). L adopts values between 0 (total water limitation of plants, i.e. no photosynthesis and growth possible) and 1 (no water limitation). For a detailed derivation of the L index, see Gerten et al. (submitted for publication). The LPJ model was run for the period 1961–1990, forced by monthly temperature, precipitation, and cloud cover from the CRU TS2.0 database (Mitchell et al., 2004; http://www.cru.uea.ac.uk/link), soil texture (FAO, 1991), and atmospheric CO2 concentrations (see McGuire et al., 2001). The climate data are disaggregated to quasi-daily values (Gerten et al.,
D. Gerten et al. / Physics and Chemistry of the Earth 30 (2005) 334–338
2004a). In order to assess how much water is appropriated for crop production, the simulated ﬁelds of transpiration were analysed as the diﬀerences between a model run that accounts for both potential natural vegetation and cropland and an alternative run that accounts for natural vegetation only. Furthermore, we interpret the spatial pattern of relative soil moisture (Wr) and related co-limitation of transpiration and photosynthesis (L) as proxies for the water limitation of potential terrestrial vegetation (in this pilot study, L was not computed for cropland).
3. Results The simulated amount of green water transpired under conditions of natural vegetation is at a maximum in the humid tropics, while close to zero in desert and icecovered regions without vegetation (Fig. 1a). Global transpiration is estimated at 41,370 km3 yr 1 (39% of total precipitation over land); total evapotranspiration reaches 63,906 km3 yr 1. The latter ﬁgure lies well within the range suggested by previous studies (61,000– 73,000 km3 yr 1; Baumgartner and Reichel, 1975; Rockstro¨m and Gordon, 2001; Gordon et al., 2003a,b), and the estimated amount of transpiration is also in line with
previous estimates from biosphere models (Levis et al., 1996). Human land cover conversion (mainly forest clearances for agricultural land) substantially decreased the green water ﬂow in the respective regions, most prominently so across Europe and the US as well as in western and south-eastern Asia (especially India). In these areas, transpiration is often lower by >100 mm yr 1 when cropland is considered compared to the simulation without cropland (Fig. 1b). Globally, simulated transpiration is reduced by 7.4%, i.e. by 3032 mm yr 1 (average 1961–1990) due to land cover change, which reﬂects the fact that 12% of the global land area are currently under some form of cultivation (Ramankutty and Foley, 1999). We also found that the decrease in transpiration as induced by land cover change became gradually stronger in the course of the past century. This clearly reﬂects the successive increase in global agricultural area, though the decreasing trend in transpiration was certainly modulated by climate (details in Gerten et al., 2004b). Main reasons for the lower transpiration from agricultural areas compared to woodland are shallower rooting depths, lower interception losses, and reduced growing seasons. The latter imply longer time periods during which land lies fallow and transpiration (green water ﬂow) is taken over by soil evaporation
Fig. 1. Long-term (1961–1990) mean global green water ﬂow (transpiration) from land to the atmosphere (mm yr 1) as simulated by the LPJ model. (a) Green water ﬂow from potential natural vegetation and (b) diﬀerence between the simulation with and the simulation without cropland and pastures.
20° 0° lat.
D. Gerten et al. / Physics and Chemistry of the Earth 30 (2005) 334–338
0 -20° -40° -60°
Fig. 2. Latitudinal breakdown of long-term annual means (1961–1990) of precipitation (in mm; grey area, plotted on right axis), and of LPJsimulated relative soil moisture, Wr (dotted line) and co-limitation of transpiration and photosynthesis, L (straight line). The simulations are for potential natural vegetation.
(white water ﬂow). Actually, the white water ﬂow (soil evaporation plus interception loss) is increased by 9.7% under human inﬂuence compared to the ÔundisturbedÕ situation, such that total evapotranspiration is decreased by 1.4% globally. These changes in evapotranspiration are manifested as a 2.2% increase in global runoﬀ (blue water ﬂow), though with considerable regional variation (Gerten et al., 2004b). These results basically agree with observations in the ﬁeld, as, for example, deforestation usually increases runoﬀ and reduces evapotranspiration (e.g. Dunn and Mackay, 1995). Soil moisture and the related water limitation of terrestrial vegetation vary considerably across the globe (Fig. 2). The spatial distribution of Wr closely follows the precipitation patterns, with a maximum in the inner tropics and a minimum in subtropical regions. L, however, is not consistently synchronised with Wr. The largest diﬀerences between the two ratios occur in mid- and high latitudes; here, Wr adopts medium values, whereas L is close to 1 (i.e. transpiration and photosynthesis are less severely limited than suggested by the soil moisture status). This divergence is attributable to the fact that in these regions, transpiration and photosynthesis are limited by temperature and radiation rather than by Wr (see Gerten et al., submitted for publication).
4. Discussion This study provides an initial assessment of green water ﬂows (plant transpiration) and their limitations as dependent on human land cover change and, implicitly, on the spatio-temporal distribution and interaction of climate, atmospheric CO2 concentration, and soil type. Two major results of the study are: (i) green water ﬂows would be notably higher in the absence of land cover change (Fig. 1), and (ii) green water ﬂows cannot be fully quantiﬁed by measures of soil moisture in isolation (see discrepancy between Wr and L, Fig. 2), as the
balance between atmospheric and soil moisture deﬁcits needs to be accounted for in a process-based manner. The consideration of the dynamics and feedbacks of all relevant processes involved in the water limitation of terrestrial ecosystems will become even more important for assessments of the future situation. For example, higher atmospheric CO2 concentration usually increases water use eﬃciency and, thus, decreases plant water limitation irrespective of the soil moisture status (Gerten et al., submitted for publication). It has to be emphasised that our results are preliminary. They may be biased due to uncertainties and biases in the underlying climate data especially in snowdominated subpolar regions. In dry regions, evapotranspiration may furthermore be underestimated due to insuﬃcient representation of human water withdrawal and of white water ﬂows from surface water (see Gerten et al., 2004a). Irrigation is not considered here, such that the human-induced decrease in green water ﬂow is probably exaggerated for regions where irrigated agriculture predominates (in particular for parts of the US and south-eastern Asia). As other studies suggest (e.g. Gordon et al., 2003a,b), the diﬀerent eﬀects of human land use change—i.e. conversion of natural into agricultural vegetation on the one hand, and irrigation on the other hand—tend to cancel out each with respect to total evapotranspiration at global scale, though the individual contributions certainly diﬀer greatly among regions. An enhanced model version of LPJ that accounts for irrigated area, irrigation requirements, and region-speciﬁc irrigation eﬃciency (as e.g. in Do¨ll, 2002) will enable consistent assessment of green and blue water ﬂows. To better represent the latter, this updated model will furthermore consider river routing, water withdrawal for industrial or domestic purposes, and water storage in surface waterbodies. It is also intended to account for management practices (e.g. rainwater harvesting, supplemental irrigation), which substantially inﬂuence the apportionment of local precipitation into green, white, and blue water ﬂows (e.g. Savenije, 2002). Notwithstanding the shortcomings of the current model version, the here deﬁned L ratio is potentially useful in the context of integrated assessments of water resources at various spatial and temporal scales. Previous global-scale studies of human water stress have usually been focused on blue water availability (e.g. Arnell, 2003), yet green water resources are certainly also relevant. We therefore aim at a comprehensive and consistent quantiﬁcation of global green and blue water ﬂows and of water limitations of terrestrial vegetation (both natural and agricultural), using newly developed water stress indicators. ‘‘Green water stress indicators’’ may be built based upon physiologically-based variables such as the L ratio, which measures the actual degree of water limitation of terrestrial plants. They may be aggregated e.g. to country level, and then combined with
D. Gerten et al. / Physics and Chemistry of the Earth 30 (2005) 334–338
established blue water stress indicators (see e.g. Vo¨ro¨smarty et al., 2000) and also with indicators of the water requirements of aquatic ecosystems (Smakhtin et al., 2004). Such a more holistic approach is particularly important in view of the potentially growing scarcity of both blue and green water in many regions and the related consequences for e.g. water availability for food production.
Acknowledgements This study was basically funded by the German Ministry for Education and Research under the German Climate Research Programme DEKLIM (project CVECA). We thank the reviewer for helpful comments on a former version of the manuscript.
References Arnell, N.W., 2003. Climate change and global water resources: SRES emissions and socio-economic scenarios. Global Environ. Change 14, 31–52. Baumgartner, A., Reichel, E., 1975. Die Weltwasserbilanz. R. Oldenbourg Verlag, Munich, Vienna. Cosgrove, W.J., Rijsberman, F.R., 2000. World Water Vision— Making Water EverybodyÕs Business. World Water Council, Earthscan, London. Do¨ll, P., 2002. Impact of climate change and variability on irrigation requirements: a global perspective. Clim. Change 54, 269– 293. Dunn, S.M., Mackay, R., 1995. Spatial variation in evapotranspiration and the inﬂuence of land use on catchment hydrology. J. Hydrol. 171, 49–73. Falkenmark, M., 1997. Meeting water requirements of an expanding world population. Phil. Trans. R. Soc. Lond. B 352, 929–936. Falkenmark, M., Rockstro¨m, J., 2004. Balancing Water for Humans and Nature. Earthscan Publications Ltd., London. FAO (Food and Agriculture Organization), 1991. The Digitized Soil Map of the World (Release 1.0). FAO, Rome. Gerten, D., Schaphoﬀ, S., Haberlandt, U., Lucht, W., Sitch, S., 2004a. Terrestrial vegetation and water balance—hydrological evaluation of a dynamic global vegetation model. J. Hydrol. 286, 249–270. Gerten, D., Bondeau, A., Hoﬀ, H., Lucht, W., Schaphoﬀ, S., Smith, P., 2004b. Assessment of ‘‘green’’ water ﬂuxes with a dynamic global vegetation model. In: Webb, B., Arnell, N., Onof, C., MacIntyre, N., Gurney, R., Kirby, C. (Eds.), Hydrology: Science and Practice for the 21st Century, vol. I. British Hydrological Society, London, pp. 29–35. Gerten, D., Schaphoﬀ, S., Lucht, W., submitted for publication. Potential future changes in water limitations of the terrestrial biosphere. Clim. Change. Gordon, L., Jo¨nsson, B.F., Johannessen, A., Steﬀen, W., Falkenmark, M., Folke, C., 2003a. Large scale redistribution of global vapour
ﬂows by deforestation and irrigation. Phil. Trans. Roy. Soc. B 358, 1973–1984. ˚ ., Steﬀen, W., Falkenmark, Gordon, L., Jo¨nsson, B.F., Johannessen, A M., Folke, C., 2003b. Large-scale redistribution of global water vapour ﬂows by deforestation and irrigation. In: Gordon, L. Land use, freshwater ﬂows and ecosystem services, Doctoral Thesis in Natural Resources Management, Department of Systems Ecology, Stockholm University. IMAGE-team, 2001. The IMAGE 2.2 implementation of the SRES scenarios. A comprehensive analysis of emissions, climate change and impacts in the 21st century. RIVM CD-ROM publication 481508018, National Institute for Public Health and the Environment, Bilthoven, The Netherlands. Levis, S., Coe, M.T., Foley, J.A., 1996. Hydrologic budget of a land surface model: a global application. J. Geophys. Res. 101 (D12), 16921–16930. Lucht, W., Prentice, I.C., Myneni, R.B., Sitch, S., Friedlingstein, P., Cramer, W., Bousquet, P., Buermann, W., Smith, B., 2002. Climatic control of the high-latitude vegetation greening trend and Pinatubo eﬀect. Science 296, 1687–1689. LÕvovich, M.I., White, G.F., 1990. Use and transformation of terrestrial water systems. In: Turner, B.L., IIII (Ed.), The Earth as Transformed by Human Action: Global and Regional Changes in the Biosphere over the Past 300 years. Cambridge University Press, Cambridge, MA, pp. 235–252. McGuire, A.D., Sitch, S., Clein, J.S., 2001. Carbon balance of the terrestrial biosphere in the twentieth century: analyses of CO2, climate and land use eﬀects with four process-based ecosystem models. Global Biogeochem. Cycles 15, 183–206. Mitchell, T.D., Carter, T.R., Jones, P.D., Hulme, M., New, M., 2004. A comprehensive set of high-resolution grids of monthly climate for Europe and the globe: the observed record (1901–2000) and 16 scenarios (2001–2100). Tyndall Centre Working Paper 55, Norwich. Ramankutty, N., Foley, J.A., 1999. Estimating historical changes in global land cover: croplands from 1700 to 1992. Global Biogeochem. Cycles 13, 997–1027. Rockstro¨m, J., Gordon, L., 2001. Assessment of green water ﬂows to sustain major biomes of the world: implications for future ecohydrological landscape management. Phys. Chem. Earth. (B) 26, 843–851. Savenije, H.H.G., 2002. The role of green water in food production in sub-Saharan Africa. FAO online document, Available from . Sitch, S., Smith, B., Prentice, I.C., Arneth, A., Bondeau, A., Cramer, W., Kaplan, J.O., Levis, S., Lucht, W., Sykes, M.T., Thonicke, K., Venevsky, S., 2003. Evaluation of ecosystem dynamics, plant geography and terrestrial carbon cycling in the LPJ dynamic global vegetation model. Global Change Biol. 9, 161–185. Smakhtin, V., Revenga, C., Do¨ll, P., 2004. A pilot global assessment of environmental water requirements and scarcity. Water Int. 29, 307– 317. Vo¨ro¨smarty, C.J., Green, P., Salisbury, J., Lammers, R.B., 2000. Global water resources: vulnerability from climate change and population growth. Science 289, 284–288. Wagner, W., Scipal, K., Pathe, C., Gerten, D., Lucht, W., Rudolf, B., 2003. Evaluation of the agreement between the ﬁrst global remotely sensed soil moisture data with model and precipitation data. J. Geophys. Res. 108, 4611. doi:10.1029/2003JD003663.