Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.
OCR for page 50
Earth Observations from Space: The First 50 Years of Scientific Achievements 6 Hydrology The global water cycle links all components of the Earth system as water moves from the atmosphere to the land and ocean and through the biosphere and cryosphere. Water plays a central role in the climate system due to its importance in the carbon cycle as water availability is a key factor for terrestrial photosynthesis. Water is also essential to the energy balance of the Earth because the rate of evaporation controls the latent heat flux. In addition, water is necessary to sustain all life; and, consequently, furthering the understanding of global hydrology is of central importance to society. Through some remarkable technological accomplishments, some important variables and processes associated with the water cycle can be retrieved now from satellite observations: water vapor (see Chapters 3 and 5), precipitation over oceans and land, snow, ice sheet mass and flow (see Chapter 7), continental groundwater storage, and sea surface temperature (see Chapter 8). The new perspective provided by satellite observations—revealing high temporal and spatial variability—has transformed the global understanding of hydrology more than any single observing platform could have done. At the same time, some important hydrologic measurements are not yet available from space, such as snow water equivalent in mountainous areas, soil moisture, and procedures to estimate evapotranspiration from remote sensing. PRECIPITATION ESTIMATES FROM THE TROPICAL RAINFALL MEASURING MISSION In orbit since 1997, the Tropical Rainfall Measuring Mission (TRMM) has transformed our ability to measure the spatial and temporal variability of rainfall in the tropics, especially over the oceans. Learning about precipitation over the oceans has catalyzed further understanding of air-sea interaction, the role of runoff to the seas in ocean circulation, and the vertical circulation of the oceans. In addition, it has led to great improvements in weather forecast skill, particularly for the southern hemisphere (see Chapter 3). TRMM’s ability to measure the height in the atmosphere where precipitation is generated provides information on the vertical distribution of release of latent heat, which in turn improves our knowledge of atmospheric circulation and climate. Moreover, TRMM has demonstrated that the technology for reliable precipitation measurements from space is now available and has provided the community with important lessons to guide the design of the Global Precipitation Mission (GPM; NRC 2007c). Before TRMM, rainfall estimates were obtained from ground-based sources (e.g., rain gauges and radar) and from satellites with visible, infrared, and passive microwave sensors (e.g., Advanced Microwave Scanning Radiometer [AMSR] for the Earth Observing System [EOS], Advanced Microwave Sounding Unit B, and Special Sensor Microwave/Imager [SSM/I]). The scientific accomplishments of TRMM are due primarily to two innovative aspects of the mission: TRMM’s complementary suite of instruments and its orbital characteristics (NRC 2006). TRMM’s instruments include a microwave imager, a visible and infrared scanner, and a lightning imaging sensor all on the same platform along with the first-ever precipitation radar in space. The suite of instruments on TRMM allows for intercalibration among the instruments as well as cross-calibration with sensors on other platforms. TRMM’s precipitation radar provides direct, fine-scale observations of precipitation and its vertical distribution. The satellite’s 35-degree inclination orbit and low altitude (402.5 km) allows for sampling well beyond the tropics to 60° N/S, but sampling in the tropics is more frequent. The orbit is not sun-synchronous; therefore, in each month it acquires measurements at all longitudes and all times of day. These advantages augment the spatial and temporal views of standard polar-orbiting environmental satellite trajectories. However, due to TRMM’s narrow swath, data for any given storm or location are available infrequently. The GPM proposes to overcome
OCR for page 51
Earth Observations from Space: The First 50 Years of Scientific Achievements most of TRMM’s limitations and is central to ensuring the availability of remotely sensed precipitation measurements for climate research (NRC 2007a). Through its technological innovations, TRMM has enabled the following scientific accomplishments for hydrology and climate: establishing rainfall climatology, quantifying the diurnal cycle of precipitation and convective intensity, and profiling latent heating (NRC 2006). TRMM data have also contributed to operational use: near-real-time TRMM-based multisatellite estimations of rainfall are being used to detect floods in the United States and especially overseas where conventional information is lacking. The National Oceanic and Atmospheric Administration’s (NOAA) National Environmental Satellite Data and Information Service uses TRMM data as part of its Tropical Rainfall Potential Program to estimate flood potential in hurricanes (Box 6.1, Figure 6.1). The National Aeronautics and Space Administration’s (NASA) TRMM-based Multisatellite Precipitation Analysis is used globally to detect floods and monitor rain for agricultural uses. The Naval Research Laboratory Monterey and the National Centers for Environmental Prediction use TRMM data as a key part of their multisatellite rain estimates. TRMM data are central to the success of these efforts because of their accuracy and the significant sampling coverage by TRMM in the tropics. The scientific accomplishments and operational advantages of TRMM have spurred the development of the GPM follow-on mission, scheduled for launch in 2013 (NRC 2007a). GPM will consist of a core spacecraft with a dual-frequency precipitation radar and a multifrequency microwave radiometric imager with high-frequency capabilities to serve as an orbiting “precipitation physics laboratory.” In addition to the core spacecraft, GPM will include a constellation of current and planned satellites with passive microwave radiometers. Together, the system will provide calibrated global precipitation at 2- to 4-hr intervals. SEASONAL SNOW COVER Of the seasonal changes that occur on Earth’s land surface, perhaps the most profound is the accumulation and melt of seasonal snow cover. Snow influences climate, weather, and the water balance. Snow cover has significant effects on energy and mass exchange between Earth’s surface and atmosphere and is an important reservoir of fresh water. Its high albedo changes the surface radiation balance; its low thermal diffusivity insulates the ground; and it is a wet, cold surface in the context of heat and moisture fluxes. Therefore, snow cover exerts a huge influence on the hydrologic cycle during the winter and spring for much of Earth’s land area. Near many mountain ranges, the seasonal snow cover is the dominating source of runoff, filling rivers and recharging aquifers that more than a billion people depend on for their water resources (Barnett et al. 2005a). Snow affects large-scale atmospheric circulation. Early-season snow cover variability in the northern hemisphere, for example, leads to altered circulation patterns, suggesting implications for climate predictability (Cohen and Entekhabi 1999). For four decades, satellite remote sensing instruments have measured snow properties. These weekly measurements represent one of the longest satellite-derived climate data records, which now enables scientists to study long-term trends in seasonal snow cover (Frei and Robinson 1999). At optical wavelengths, sensors such as the NOAA Advanced Very High Resolution Radiometer (AVHRR) and the Landsat Thematic Mapper (TM) have been used to produce maps of snow cover at both continental and drainage-basin scales. In the EOS era, snow-cover products are available from the Moderate Resolution Imaging Spectroradiometer (MODIS), the Multiangle Imaging Spectroradimeter (MISR), and the Advanced Spaceborne Thermal Emission Reflection and Radiometer (ASTER). Snow-water equivalent (the depth of liquid water that the snowpack would produce if it melted) is regularly estimated at coarse spatial resolution from passive microwave data, including SSM/R, SSM/I, and the EOS instrument AMSR-E in a time series that goes back to 1978. However, at finer spatial resolution, necessary for the mountains, measuring snow-water equivalent is a difficult problem; and a proposed sensor for Snow and Cold Land Processes (SCLP) is recommended as one of 17 high-priority missions for launch before 2020 (NRC 2007a). König et al. (2001) and Dozier and Painter (2004) have reviewed developments in remote sensing of snow and ice. Among them is the use of snow-covered area from MODIS in hydrologic analysis and modeling (Box 6.2, Figure 6.2). Through updates of a runoff model with measurements of snow cover, seasonal streamflow forecasts have been improved (McGuire et al. 2006). Unlike surface measurements, satellite observations are able to show the distribution of snow over the topography, revealing that considerable snow at higher elevations remains after all snow has disappeared from the surface measurement stations. An additional property measured from MODIS is snow albedo. In the current generation of climate and snowmelt models, snow albedo is typically either prescribed or represented by empirical aging functions, when truly it is a dynamic variable affected by grain growth and light-absorbing impurities. Newer analyses of snow cover are incorporating the seasonal evolution of both the snow cover and its albedo. In the visible part of the spectrum, clean, deep snow is bright and white, irrespective of the size of the grains. Beyond the visible wavelengths in the near infrared and shortwave infrared, however, snow is one of the most “colorful” substances in nature. Newly fallen snow usually has a fine grain size, but metamorphism and sintering throughout the winter and spring increase the grain size, bond grains together, and reduce reflectance in wavelengths beyond about 0.8 μm (Warren 1982). This behavior of snow is important to the snowpack’s energy balance because the decrease in albedo often occurs during the spring when the
OCR for page 52
Earth Observations from Space: The First 50 Years of Scientific Achievements BOX 6.1 Improved Understanding of Hydrology and Climate from TRMM TRMM-based multisatellite data are being used as input into hydrologic models, including the Land Data Assimilation System, to better understand land-atmosphere interactions on scales of days to years (Rodell et al. 2004) and to study variations in river runoff (Fekete et al. 2004). These same data are being used to monitor crops in Central America and elsewhere and as input into river forecast models in South Asia and other locations. Analysis of TRMM precipitation radar data has been used to discover orographic precipitation processes and diurnal cycles of rainfall causing flash floods in headwater streams (Barros et al. 2004). TRMM observations led to the discovery of extremely tall convective towers within the vertical precipitation profiles of tropical cyclones. Kelley et al. (2004) reported that the chance of intensification increases when one or more of these “hot towers” exist in the tropical cyclone’s eyewall (Figure 6.1). FIGURE 6.1 A cross-sectional view of Hurricane Katrina through the eye of the storm, as observed from TRMM. This image shows the horizontal distribution of rain intensity on August 28, 2005, when Katrina was a Category 3 hurricane with maximum sustained winds of 100 knots (115 mph). Rain rates in the central portion of the swath are from the TRMM precipitation radar, and the rain rates in the outer swath are from the TRMM microwave imager. The rain rates are overlaid on infrared data from the TRMM visible infrared scanner. Two isolated hot towers (in red) are visible: one in an outer rain band and the other in the northeastern part of the eyewall. The height of the eyewall tower is 16 km. Towers of this height near the core are often an indication of intensification as was true with Katrina, which became a Category 4 storm soon after this image was taken. SOURCE: NASA (2005).
OCR for page 53
Earth Observations from Space: The First 50 Years of Scientific Achievements BOX 6.2 High-Resolution Seasonal Snow Cover Data Improve Climate and Hydrology Models Because of the influence of seasonal snow cover on climate, weather, and water balance, it is a crucial quantity for climate and hydrology models. Furthermore, daily maps are necessary for hydrologic and climate models due to the dynamic nature of snow cover, which changes at a slower timescale than atmospheric phenomena but faster than other surface covers. The availability of daily global observations of this parameter was inconceivable prior to the satellite era. Nowadays, the global MODIS snow-cover product (Hall et al. 2002) is produced daily and as an 8-day composite at 500-m spatial resolution. For global climate models, daily snow cover is produced at 0.05° latitude-longitude grid cells (about 5.5 km in the north-south direction) along with monthly global composites. The composites are necessary because cloud cover and viewing geometry affect the daily images (Figure 6.2). FIGURE 6.2 MODIS image (left) and interpreted snow (white) and cloud (pink) cover over the Sierra Nevada, January 5, 2003. SOURCE:SOURCE: http://modis-snow-ice.gsfc.nasa.gov/images.html. incoming solar radiation becomes greater as the solar elevation increases and the days get longer. In the context of hydrologic models, this albedo decay has spatial variability. Molotch et al. (2004) examined snow ablation from a grid-based distributed snowmelt model, using field data from extensive snow surveys during the melt season to initialize the model with a spatial distribution of snow-water equivalent and then to test the model with subsequent surveys. Remotely sensed albedo typically differed by 20 percent from albedo estimated using a common snow age-based empirical relation applied uniformly across the domain. Snowpack models are just beginning to incorporate albedo evolution, based on the movement of water molecules in the snow to reduce the surface area of the grains in comparison to their volume (Flanner and Zender 2006). A recent development in mapping snow cover and its albedo is “subpixel” analysis. Snow-covered area in mountainous terrain usually varies at a spatial scale finer than that of the ground instantaneous field of view of the remote sensing instrument. This spatial heterogeneity poses a “mixed-pixel” problem because the sensor may measure radiance reflected from snow, rock, soil, and vegetation. To use the snow characteristics in hydrologic models, snow must be mapped at subpixel resolution in order to accurately
OCR for page 54
Earth Observations from Space: The First 50 Years of Scientific Achievements represent its spatial distribution; otherwise, systematic errors may result. For example, especially in drier years, much of the snow cover is patchy at the lower elevations. An image classification that identifies each pixel as either snow covered or not may miss much of this snow. Mapping of surface constituents at subpixel scale uses a technique called “spectral mixture analysis,” based on the assumption that the radiance measured at the sensor is a linear combination of radiances reflected from individual surfaces (Figure 6.3). Snow does not have a unique reflectance in each wavelength band, but given its physical characteristics such as grain size and amount and composition of impurities, a snow end member can be chosen that results in the lowest error in the solution of the simultaneous equations (Painter et al. 2003). The information thereby derived is the fractional snow-covered area for each pixel and the albedo of that snow. DISCOVERY OF ANCIENT BURIED RIVER CHANNELS In 1981 the first Shuttle Imaging Radar (SIR-A) was launched on the space shuttle Columbia, assembled partly with spare parts from the 1978 Seasat synthetic aperture radar (SAR). With just a single frequency and one polarization, and capable of acquiring imagery at only one angle, SIR-A showed that in the dry Sahara Desert it could penetrate as deeply as 3 m. These early images from the dunes and drift sand of the eastern Sahara showed previously unknown buried valleys, geologic structures, and possible Stone Age occupation sites (McCauley et al. 1982). Radar responses from bedrock and gravel surfaces beneath wind-blown sand several meters thick delineated sand- and alluvium-filled valleys, some nearly as wide as the Nile Valley and perhaps as old as middle Tertiary. The now-vanished major river systems that carved these large valleys probably accomplished most of the erosional stripping of this extraordinarily flat, arid region. Stone Age artifacts associated with soils in the alluvium suggested areas that may have been sites of early human occupation. The presence of old drainage networks beneath the sand (Figure 6.4) provided a geologic explanation for the locations of many playas and present-day oases that have been centers of episodic human habitation. The success of the mission paved the way for a follow-on, the SIR-B in 1984, which could collect data at more than one angle by mechanically tilting its antenna, and then the SIR-C (SIR-C/X-SAR) in April and October 1994. The synthetic aperture radar on board SIR-C was fully polarimetric, capable of both transmitting and collecting information at FIGURE 6.3 Fractional snow cover over the Sierra Nevada on April 1 (left) and May 1, 2005. Total snow-covered area is 23,100 km2 in April and 14,900 km2 in May. SOURCE: J. Dozier, University of California, Santa Barbara.
OCR for page 55
Earth Observations from Space: The First 50 Years of Scientific Achievements FIGURE 6.4 Image from Shuttle Imaging Radar-A (SIR-A) showing buried river channels in the Sahara Desert. SOURCE: http://www.jpl.nasa.gov/history/hires/1981/SIR-A_image.jpg. vertical or horizontal polarizations. In addition, the antenna was electronically steerable and operated at three frequencies (1.4, 5.6, and 10 GHz). The two flights allowed investigation into the radar’s response to seasonal changes. The multiparameter images were combined and enhanced to produce some of the most spectacular radar images ever seen. ANALYSIS OF GROUNDWATER FROM GRAVITY DATA A pair of satellites launched in 2002 makes up NASA’s Gravity Recovery and Climate Experiment (GRACE). The main source of variation in Earth’s gravity field is the movement of water between its three main reservoirs: the ocean, ice sheets, and groundwater. Unlike most satellite remote sensors, which measure electromagnetic radiation reflected or emitted from Earth’s surface and atmosphere, GRACE measures the distance between its two spacecraft, which changes in response to variations in Earth’s mass—and therefore gravity—on the surface below them. The GRACE measurement also senses mass change within the Earth—a capability demonstrated by the measurement of seasonal change in continental aquifers. Wahr et al. (2004) use GRACE data to compare groundwater storage with a hydrologic model in the Mississippi and Amazon River basins and in the drainage flowing into the Bay of Bengal (Figure 6.5). When averaged over 1,000 km or more, the mass estimates inferred from the GRACE data clearly show annually varying changes in continental water storage, along with seasonal variability in the amount of water in the ocean. The amplitudes and phases of those signals are in general agreement with the hydrologic model. The inferred mass signals over the ocean are consistent with estimates of the water stored in the groundwater. Although the agreement degrades with decreasing averaging radius, the largest water storage signals are still clearly evident at averaging radii as short as 400 km. The globally averaged uncertainty in the amplitude of the annually varying mass signal recovered from these GRACE fields is 1.0 cm for a 1,000-km radius. USE OF SATELLITE-DERIVED ELEVATION DATA IN HYDROLOGY In February 2000, with the aid of a 60-m (200-ft) boom added to the SIR-C, the Shuttle Radar Topography Mission (SRTM) circled Earth for 10 days mapping 80 percent of the world’s land area. The resulting high-resolution topographic map is the most accurate available and constitutes one of the major accomplishments of the nation’s space program. SRTM (Farr et al. 2007) provides a worldwide topographic data set between 60° N and S latitudes with a consistent datum. Many areas otherwise lack topographic data, so these data enable spatial hydrologic modeling that would otherwise be impossible. Figure 6.6 shows an elevation and relief map of the whole African continent. Since its launch, digital elevation models created from SRTM have been used in many applications, most notably tectonics, geomorphology, and hydrology. Because of their global consistency, SRTM data link continental hydrology with the oceans. In hydrologic investigations, the first information in characterizing a problem is often the topography of a drainage basin. From the elevations, slopes and aspects can be estimated, which are essential for calculations of solar and longwave radiation that can be used in spatially distributed energy balance models of snowmelt (Cline et al. 1998), photosynthesis, and evapotranspiration (Anderson et al. 2003). Analytical software packages use these slopes and aspects as input parameters to delineate drainage basin boundaries, to characterize basins for their distribution of slopes, and in routing water from precipitation or snowmelt (Tarboton 1997). An additional hydrologic application of SRTM data has been to measure water surface elevations directly (Alsdorf et al. 2007), which contributes to the improvement of flood forecasting. Providing accurate flood forecasts from satellite observations is a high-priority mission with the potential to save lives and property (NRC 2007a). This important societal challenge
OCR for page 56
Earth Observations from Space: The First 50 Years of Scientific Achievements FIGURE 6.5 The mass variability within (a) the Mississippi River basin, (b) the Amazon River basin, and (c) a drainage system flowing into the Bay of Bengal, as inferred from the GRACE measurements (dots). Also shown are results inferred from a hydrologic model, as well as the best-fitting annual signal for both the GRACE values and the model predictions. Bottom panels show the optimal averaging kernels used to recover this mass variability. SOURCE: Wahr et al. (2004). Reprinted with permission from the American Geophysical Union, copyright 2004. cannot be answered adequately with the current global in situ networks designed to observe river discharge. Knowledge of soil moisture (and snow water storage, where relevant), surface water area, the elevation and slope of the water surface, and accurate hydrologic models is required to meet this challenge. Although attempts to estimate soil moisture from the AMSR-E sensor have been made, they are only at the early experimental stage. Nevertheless, results have shown promises and the proposed Soil Moisture Active-Passive mission is central to making progress toward reliable flood hazard assessments (NRC 2007a). Despite the many accomplishments highlighted in this chapter, important challenges remain such as the GPM, soil moisture estimates, surface water and ocean topography (to improve estimates of water stored in lakes, reservoirs, wetlands, and rivers), and improved estimates of snowpacks (NRC 2007a).
OCR for page 57
Earth Observations from Space: The First 50 Years of Scientific Achievements FIGURE 6.6 Elevation and relief map of Africa from the Shuttle Radar Topography Mission. Color coding is directly related to topographic height, with brown and yellow at the lower elevations, rising through green, to white at the highest elevations. Blue areas on the map represent water within the mapped tiles, each of which includes shorelines or islands. SOURCE: NASA Jet Propulsion Laboratory.