Wetlands' Outsize Influence on Climate
Wetlands occupy only a few percent of Earth's surface. However, their influence on climate is much larger than their small area suggests due to the important role they play in the world's water cycle and because they are the world's largest source of methane, a powerful greenhouse gas. Consequently, understanding and ultimately predicting flooding and associated biogeochemical activity in wetlands is a high priority in climate research.
Water temporarily stored in flooded wetlands alters the amount and seasonality of river flow, modifies regional temperatures (cools in warm weather, warms in cold weather), and humidifies surrounding areas by serving as a source for evaporation. Flooded and non-flooded wetlands, including irrigated rice fields, account for about 30-40% of methane emitted from Earth's surface to the atmosphere each year and are likely to respond to changing climate in coming years.
Methane is produced when organic material in soils is broken down by microbes called methanogens (methane producers) in the absence of oxygen. When water fills spaces among soil particles, as it does in wet and flooded soils, oxygen is in short supply and methanogens thrive. Methane is produced in sediments below the soil's water table and travels upward through the soil, through the stems of some plants or by bubbling through standing water (ebullition). (If you've ever produced bubbles by disturbing the bottom of a pond with a stick, you've seen methane ebullition.) Methane emissions from wetlands are large and respond to annual fluctuations in weather; more methane is released when it's warm and wet, less when it's cool and dry. Inundation dynamics of wetlands, together with temperature, are the primary controllers of large interannual variations in the amount of methane added to the atmosphere.
Figure 1 (at right): Images of varied wetland terrains. (a) mixed landscape of sparse shrubs, evergreens, and temporary water in Alaska (Credit: Elaine Matthews, NASA/GISS); (b) fen-dominated landscape in Kejimkujik National Park, Nova Scotia (Credit: Clayton Rubec, Environment Canada); (c) patterned string fen in Newfoundland Labrador (Credit: Doyle Wells, Forestry Canada); (d) cypress swamp in southern Mississippi's Natchez Trace Parkway (Credit: U.S. National Parks Service). (e) Victoria regia water lilies in Amazonia (Credit: UNEP World Conservation Monitoring Centre); and (f) fishermen in the Al-Hawizeh reed marshes of southern Iraq (Credit: U.S. Army Corps of Engineers). Click on any image to see a larger version.
Natural wetlands are found from the tropics to the arctic but are difficult to characterize due to their exceptional variability. Figure 1 illustrates this variety for a collection of wetland complexes. For example, some wetlands in northern Russia, Alaska and Canada have very wet soils and/or temporary pools with a sparse cover of bushes and evergreen trees (Fig. 1a); other boreal (northern) wetlands are composed of flooded reed communities adjacent to rivers (Fig. 1b) or string bogs on sloping lands made up of shallow elongated pools interwoven with peat-rich ridges sometimes populated by small trees (Fig. 1c). Temperate wetlands are often tree-dominated swamps like the cypress swamps found in the southeastern United States. (Fig. 1d). Tropical and subtropical wetlands include seasonally-flooded forests and grasslands along rivers such as the Congo and Niger Rivers (west Africa) and the Amazon River (South America) (Fig. 1e) and seasonally-flooded swamps dominated by tall reeds in the Middle East and North Africa (Fig. 1f).
This wide spectrum of vegetation cover, hydrological regime and natural seasonality means that defining wetlands and/or inundation is not straightforward and no overall consensus on the subject exists. In addition, researchers studying lakes, rivers and wetlands tend to work independently. In reality, meaningful distinctions among wetlands, rivers and lakes may not be possible in some regions and/or seasons which complicates the development of techniques to monitor these environments.
The importance of wetlands, rivers and lakes in hydrological and biogeochemical cycles has been well known for quite awhile. Despite the publication of many local, and a few regional, studies of wetlands using satellite data, characterizing and quantifying large-scale distributions and interannual variations of wetlands, and particularly flooding, has been a difficult challenge. One major obstacle has been that the multiple satellite instruments needed for the job were not all operating until relatively recently. Even after satellite observations became available, substantial work was needed to determine the combination of data that best captures flooding and to develop global analysis techniques to extract the maximum information about wetlands. Research by NASA/GISS scientists and affiliated researchers in optimizing satellite observations and methodologies to quantify and monitor inundated wetlands globally recently began to yield very promising results (Prigent et al. 2001). A new paper by our research group (Prigent et al. 2007) presented and evaluated the first global, multi-year (1993-2000) estimate of monthly inundation extent.
Research results were derived from a multi-satellite method employing instrumental data with complementary strengths: 1) passive microwave land-surface emissivities calculated from the Special Sensor Microwave/Imager (SSM/I) and International Cloud Climatology Project (ISCCP) observations, 2) the active microwave response from the European Remote Sensing (ERS) satellite scatterometer responses, and 3) the Advanced Very High Resolution Radiometer (AVHRR) visible and near-infrared reflectances and the derived Normalized Difference Vegetation Index (NDVI) calculated from them. The data were used to calculate global, monthly inundated fractions of equal-area grid cells, about 25 km2, taking into account the contribution of vegetation to the passive microwave signal.
Figure 2 illustrates the global results of this study. The upper panel shows the typical duration, in months, of inundation for the 1993-2000 period; the lower panel shows mean maximum inundated fraction for the same period. Maximum global inundated area (August) averages 5.86×106 km2 for 1993-2000, while the minimum area (December) averages 2.12×106 km2. Maximum inundated fraction and longest period of flooding generally coincide (Bangladesh, southeast Asia, the Amazon and Parana Rivers in South America, and the Niger River in equatorial Africa).
Figure 2 (above): Upper panel: mean annual number of inundated months for 1993-2000. Lower panel: Mean maximum fractional inundation. Both mapped at ~25 km2 equal area grid equal to 773 km2. (View as large GIF or PDF) (Credit: NASA/GISS and CNRS/Observatoire de Paris)
Initial evaluation of the duration and seasonality of flooding against independent climate data and satellite observations over boreal and tropical regions confirms the large-scale realism of the retrievals, but validating the new results is particularly difficult due to the scarcity of information on the distribution and dynamics of large-scale inundation. We are working with colleagues at the Jet Propulsion Laboratory and elsewhere to compare high resolution (100 m) radar-derived water and wetland classifications to assess in more detail the performance of the multi-satellite approach for a suite a representative ecosystems in addition to combing the literature and libraries of satellite imagery for reports, maps, and images of wetlands and lakes to assess the new results.
This unique dataset of global inundation dynamics is crucial to various applications including hydrological and methane modeling, water management, and climate modeling. We are particularly interested in investigating possible recent trends in inundation dynamics and their interactions with the global methane cycle.
Prigent, C., E. Matthews, F. Aires, and W. B. Rossow, 2001. Remote sensing of global wetland dynamics with multiple satellite data sets. Geophys. Res. Lett., 28, 4631-4634.
Prigent, C., F. Papa, F. Aires, W.B. Rossow, and E. Matthews, 2007. Global inundation dynamics inferred from multiple satellite observations over a decade. J. Geophys. Res., 112, D12107, doi:10.1029/2006JD0078472.
Please address all inquiries about this research to Elaine Matthews.