www.elsevier.com/locate/sna
Spectral Signatures of Coral Reefs: Features from Space Dan Lubin,* Wei Li,† Phillip Dustan,‡ Charles H. Mazel,§ and Knut Stamnes†
The spectral signatures of coral reefs and related scenes,
as they would be measured above the Earth’s atmosphere, are calculated using a coupled atmosphere-ocean discrete ordinates radiative transfer model. Actual measured reflectance spectra from field work are used as input data. Four coral species are considered, to survey the natural range of coral reflectance: Montastrea cavernosa, Acropora palmata, Dichocoenia stokesii, and Siderastrea siderea. Four noncoral objects associated with reefs are also considered: sand, coralline algae, green macroalgae, and algal turf. The reflectance spectra as would be measured at the top of the atmosphere are substantially different from the in situ spectra, due to differential attenuation by the water column and, most importantly, by atmospheric Rayleigh scattering. The result is that many of the spectral features that can be used to distinguish coral species from their surroundings or from one another, which have been used successfully with surface or aircraft data, would be obscured in spectral measurements from a spacecraft. However, above the atmosphere, the radiance contrasts between most coral species and most brighter noncoral objects remain noticeable for water column depths up to 20 m. Over many spectral intervals, the reflectance from dark coral under shallow water is smaller than that of deep water. The maximum top-ofatmosphere radiances, and maximum contrasts between scene types, occur between 400 nm and 600 nm. This study supports the conclusions of recent satellite reef mapping exercises, suggesting that coral reef identification should
* Scripps Institution of Oceanography University of California, San Diego, La Jolla, California † Geophysical Institute, University of Alaska, Fairbanks, Alaska ‡ College of Charleston, Charleston, South Carolina § Physical Sciences, Inc., Andover, Massachusetts Address correspondence to D. Lubin, Scripps Inst. of Oceanography, Univ. of California, San Diego, 9500 Gilman Dr. La Jolla, CA 920930221. E-mail:
[email protected] Received 27 May 1999; revised 11 January 2000. REMOTE SENS. ENVIRON. 75:127–137 (2001) Elsevier Science Inc., 2001 655 Avenue of the Americas, New York, NY 10010
be feasible using satellite remote sensing, but that detailed reef mapping (e.g., species identification) may be more difficult. Elsevier Science Inc., 2001
INTRODUCTION The world’s coral reefs are highly susceptible to damage by a variety of human activities, and by global climate change. Studies of individual coral reef ecosystems have reported catastrophic damage by either anthropogenic pollution or the water column or by physical impacts such as ship groundings (Dustan and Halas, 1987; Hughes, 1994; Hatziolos et al., 1998). It is not an exaggeration to suggest that the world’s coral ecosystems could, through neglect, suffer severe degradation leading to near-extinction and ecosystem collapse within a few decades. To date there has been no successful program to produce a complete global map of the world’s coral reefs. Of the known reef ecosystems, some are very well studied by ongoing field programs while the health of others remains unmonitored due to their remoteness. To map the global distribution of coral reef ecosystems, and to monitor the growth or deterioration of coral reefs worldwide, satellite remote sensing will be required in some capacity. The Landsat and Systeme Probatoire de l’Observations de la Terre (SPOT) multispectral shortwave imagers offer some potential for coral reef identification (Bour and Pichon, 1997; Holden and Ledrew, 1999; Miller and Cruise, 1995; Mumby et al., 1998). For example, Mumby et al. (1998) tested a retrieval algorithm using the Landsat Thematic Mapper (TM), comparing the satellite-based estimates of coral reef extent against similar 1-m-resolution retrievals from an airborne imaging spectrometer; the satellite-based method yielded an overall accuracy of 75%. In this study, we use a coupled ocean-atmosphere radiative transfer model to further investigate the potential of satellite remote sensing, by examining how the spectral and 0034-4257/00/$–see front matter PII S0034-4257(00)00161-9
128
Lubin et al.
radiometric signatures of coral reef objects should appear above from low earth orbit after being modified by water column and atmospheric attenuation. For example, it is well known that sand has a considerably higher reflectance at all visible wavelengths than any living coral. Does this significant reflectance difference that we can measure in situ manifest itself in suitable radiance contrast at the top of the atmosphere? If so, is this also true for other noncoral objects (e.g., algal turf) that are more reflective than living coral, but not as reflective as sand? We can also investigate the spectral reflectances of coral reef objects as they would appear above the atmosphere. Holden and Ledrew (1999) have shown i) that there are robust differences between in situ reflectance spectra of living corals and those of related noncoral objects, ii) that these differences are independent of geographic sampling and of coral morphology, and iii) that these spectral contrasts can be used to construct a scene identification algorithm based on differential reflectance. Does this type of differential reflectance spectroscopy remain applicable to scenes observed from above the atmosphere? It is known that differential absorption by the water column will modify considerably the spectral reflectance of an object at depth (Lyzenga, 1981; Maritorena et al., 1994), but what is the combined effect of the water column and the atmosphere? Finally, we can estimate both the spectral radiances and reflectances that would be measured by a satellite instrument. This gives us the ability to i) make a first-order estimate of what spatial resolution might be available using standard CCD technology with spectral bandwidths optimized for coral reef studies and ii) discuss the suitability of existing sensors such as Landsat TM, which were originally designed for terrestial applications, for measuring the radiometric signatures of coral reef objects. RADIATIVE TRANSFER MODEL AND INPUT DATA We base this study on the measured reflectance spectra of eight objects or scene types: sand, coralline (red crustose) algae, green macroalgae, algal turf, and four different coral species. It is obviously necessary to consider sand, if we are discussing the potential for distinguishing coral reefs from their surroundings. The coralline algae, green macroalgae, and algal turf reflectance spectra are considered because they represent components of the shallow water reef community. Montastrea cavernosa and Acropora palmata are chosen as two coral species having very different spectral reflectances (brighter and darker than average corals, respectively). Dichocoenia stokesii and Siderastrea siderea are chosen because they are intermediate in brightness at all visible wavelenghts between M. cavernosa and A. palmata. These reflectances, shown in Figure 1, were measured relative to spectralon using an underwater spectrometer (Mazel, 1997). The study site was a small Bahamian reef patch, Rainbow Gardens, located on the lee side
Figure 1. Spectral reflectances of the eight scene types considered in this study, measured by an underwater spectrometer. Reflectance shown is relative to spectralon, A) for the noncoral objects and B) for the living corals.
of Iguana Cay, a small island approximately 3 km north of Lee Stocking Island (23⬚N, 76⬚W), Exumas, Bahamas. The coral community structure and distribution has been described by King (1995). Each reflectance spectrum shown in Figure 1 is an average of a small set of sample measurements. We note that our sample spectrum of green macroalgae has a reflectance that is similar to that of dark brown macroalgae reported by Maritorena et al. (1994). The conclusions of this study are related to both the wide variability and continuous gradient in spectral reflectances of objects related to coral reefs, and do not require consideration of the standard deviations of each sample set in our radiative transfer modeling. To examine how these scene types might appear from space, after light reflected from them has been further attenuated by the water column and scattered by the atmosphere, we utilize a discrete ordinates radiative transfer algorithm (Stamnes et al., 1988). Lyzenga (1981) and Maritorena et al. (1994) have shown how the water column will modify the apparent reflectance spectrum of an object on
Spectral Signatures of Coral Reefs: Features from Space
Figure 2. Diagram of the important features of the coupled ocean-atmosphere discrete-ordinates radiative transfer model. The atmosphere, modeled using 35 layers each of uniform optical depth, has total optical depth sa; sa is the sum of Rayleigh scattering, aerosol scattering/absorption, and ozone absorption optical depths, and is a function of wavelength. The water column is modeled as a single layer having total optical depth sw, which is also a function of wavelength as specified by the Smith and Baker (1981) attenuation coefficients. The extraterrestrial solar flux is l0F0, where l0 is the cosine of the solar zenith angle. At the air-water interface, the direct solar beam is refracted into the angle l0w. As light from all downward directions (2p steradians) crosses the air-water interface, it is refracted into a cone of less than 2p steradians (region II). In the water column, light scattered outside this cone (region I) cannot reach the atmosphere. The reflectance of the reef floor is approximated as being Lambertian (isotropic).
129
ance calculation within the water column. Jin and Stamnes (1994) have derived a method that adds the appropriate number of quadrature points to the water column, given a desired number of quadrature points in the atmosphere. This method includes both scattering and absorption by both media, and allows the vertical structure of both media to be specified by any number of horizontally homogeneous layers. As input data to this radiative transfer model, we use the measured reflectance spectra of Figure 2 as the lower boundary condition, specifying a Lambertian surface albedo. The Lambertian surface albedo is a modeling assumption, as the scene bidirectional reflectance distribution functions were not measured by the underwater spectrometer. The water column is treated as a single layer whose thickness (depth) we can vary. Spectral attenuation coefficients in the water column are taken from Smith and Baker (1981) for clear ocean waters. The radiative transfer model requires that we assume a flat air-water interface, and neglect effects of ocean surface roughness at present. The vertical structure and composition of the atmosphere is specified by the tropical model atmosphere in LOWTRAN 7 (Kneiszys et al., 1988). Tropospheric aerosols are included, as specified by the maritime background model in LOWTRAN 7. The ozone column abundance (relevant to visible wavelengths via the Chappuis absorption bands) is set at 250 Dobson units, which is appropriate for tropical latitudes year round. The radiance calculations are carried out with 16 (atmospheric) computational streams for 4p steradians, with the radiances solved for each of the 16 Gaussian quadrature angles (Chandrasekhar, 1960). For clarity, we discuss radiances at the Gaussian quadrature angle closest to the downward-looking direction (11.4⬚ off nadir), as this angle is the most relevant to a potential satellite instrument operating with high spatial resolution. SPECTRAL SIGNATURES OF CORAL REEFS
the bottom, and how this effect can be treated in remote sensing retrievals. Here we are interested in both this spectral transformation and in estimates of the actual radiance at the top of the atmosphere. Therefore, for this study we use a discrete ordinates radiative transfer formulation by Jin and Stamnes (1994), to solve for the top-of-atmosphere radiances directly (Fig. 2). This formulation allows for numerical solution of the radiative transfer equation in a medium where the index of refraction changes (i.e., at the air-water interface). As light propagating through the atmosphere in all downward directions (2p steradians) crosses the air-water interface, it is refracted into a cone of less than 2p steradians (angular region II in Fig. 2). In the water column, light from within this cone can be scattered to and from the region outside it, but light scattered outside this cone (angular region I in Fig. 2) cannot reach the atmosphere. Therefore, a complete radiative transfer solution for a coupled atmosphere-ocean system requires that additional computational streams be added for the radi-
We first examine the upwelling spectral radiance at the air-water interface, when the water column depth is five meters. Figure 3 shows this upwelling radiance, at viewing angle 11.4⬚, for all eight objects. We see that the useful signal for remote sensing lies between 400 nm and 600 nm. The maxima of in situ scene reflectances (Fig. 1) for longer wavelengths are offset by the strong water column attenuation, such that the overall radiances for wavelengths longer than 600 nm are very small (⬍0.01 W m⫺2 sr⫺1 nm), and the radiance contrasts between scene types are poor. If we examine the upwelling radiance at the top of the atmosphere (Fig. 4), we see how the water-leaving spectral radiance is modulated by the atmospheric Rayleigh scattering. For wavelengths between 400 nm and 500 nm, the top-ofatmosphere radiance is typically a factor of 2 or more larger than the water-leaving radiance, due to Rayleigh scattering being the strongest at the shorter wavelengths. Between 400 nm and 600 nm, some scene types remain distinct from one another. The sand and coralline algae are the brightest
130
Lubin et al.
Figure 3. Upwelling spectral radiance at the air-water interface, in the near-nadir direction (polar angle 11.4⬚), for A) the noncoral objects and B) the coral species. The water column depth is 5 m. The solar zenith angle is 30⬚.
Figure 4. The spectral radiances at the top of the Earth’s atmosphere, for a solar zenith angle of 30⬚, A) noncoral objects, B) living corals. The water column depth is 5 m.
Figure 5 shows the top-of-atmosphere spectral reflectance, defined as Eq. (1): scenes, with maximum spectral radiances of 0.145 W m⫺2 sr⫺1 nm and 0.100 W m⫺2 sr⫺1 nm, respectively; while all four coral species considered here are noticeably darker than coralline algae and yield top-of-atmosphere radiances nearly a factor of 2 smaller than that of sand. However, the two other noncoral objects yield top-of-atmosphere radiances that are much closer to those from living corals. Spectral radiances from the algal turf are typically only 20% larger than those from the brightest coral (M. cavernosa), and spectral radiances from the green macroalgae lie squarely within the range of those from living coral. Thus, we see that the natural variability in spectral reflectance of coral reef objects yields a continuous gradient in radiance as measured from above the atmosphere, and that living corals and noncoral objects are not always distinct from one another. Although sand should be readily distinguishable from other objects, there may always be some ambiguity in resolving some types of algae from living coral with a satellite instrument.
Rk(h0, h, u)⫽100pIk(h0, h, u)/Fk(h0),
(1)
where Fk(h0) is the extraterrestrial solar spectral irradiance. The reflectance defined this way is not the true reflectance (albedo) of the ocean-atmosphere system (which would be the ratio of upwelling flux to extraterrestrial solar flux), but is instead a convenient and often used convention for scaling radiances when displaying satellite imagery. For a nadir or near-nadir view, this scaled radiance is actually smaller than the true albedo, due to the angular redistribution of radiation that favors the direction of the solar beam (specular reflection, e.g., Lubin and Weber, 1995). When the calculations are examined with this scaling, it is apparent that the reflectance contrasts between sand, coralline algae, and the corals are large enough to be practical for remote sensing; however, this is not true for green macroalgae, algal turf, and the corals. Furthermore, the effect of atmospheric Rayleigh scattering is to render the slopes of the top-of-atmosphere reflectance spectra between 400 nm and 600 nm very similar for most of the objects (coralline
Spectral Signatures of Coral Reefs: Features from Space
Figure 5. The spectral reflecance, or scaled radiance defined as 100pIk(h0, h, u)/Fk(h0), at the top of the Earth’s atmosphere, for a solar zenith angle of 30⬚, for A) noncoral objects, B) living corals. The water column depth is 5 m.
algae, green macroalgae, algal turf, A. palmata, D. stokesii, S. siderea). The exceptions are sand, with its higher overall reflectance, and M. cavernosa between 450 nm and 480 nm. Thus, many of the objects’ unique spectral features that are evident in situ or at the air-water interface (Figs. 1 and 3) are obscured by the atmospheric column. Following principles used in ocean color remote sensing, it should be possible to use a radiative transfer model to remove atmospheric effects from a satellite measurement and retrieve the ocean surface reflectance. However, for this to be successful, the satellite sensor must have enough sensitivity and precision to detect the spectral differences between coral reef objects after they have propagated through the entire atmosphere column. If these topof-atmosphere differences are smaller than the resolution or precision of the sensor, then removing the atmospheric effects will yield no information. In Figure 6, the upwelling spectral radiances at the top of the atmosphere are shown as a function of water column depth. Figure 7 shows the corresponding upwelling
131
spectral radiances at the ocean surface. Between 500 nm and 600 nm for all objects, the upwelling radiance decreases noticeably with increasing water depth as the water column attenuates the bottom-reflected radiance component. For the four coral species, the radiance changes very little with water depth for wavelengths shorter than 450 nm, and at these shorter wavelengths the largest water-leaving radiances occur for water column depth 20 m. At these shorter wavelengths, the corals have a reflectance that is less than the reflectance of the model’s optically deep water. At a depth of 20 m, there is very little difference in radiance between the four scene types around 550 nm (top-of-atmosphere radiances generally less than 0.04 W m⫺2 sr⫺1 nm), signifying that most of the bottom-reflected radiance is attenuated by the water column at this depth. For sand and coralline algae, there is substantial variability in radiance with water column depth for wavelengths between 500 nm and 600 nm, which we would expect as these scene types have higher intrinsic reflectances which would contribute to the total space-measured radiance over a greater range in water column depth. The intrinsic reflectances of the M. cavernosa and algal turf are large enough that the top-of-atmosphere radiances, between 500 nm and 600 nm, over water column depths 1–3 m are twice as large as the radiances over water column depth 20 m (essentially infinite ocean depth). A. palmata and green macroalgae, on the other hand, provide such a small bottom reflectance component that the only appreciable variability in topof-atmosphere radiance with water column depth occurs around 560–580 nm. Coral objects this dark will be the easiest to distinguish from surrounding sand, but the ambiguity between green macroalgae and living corals remains apparent at all depths. UTILITY OF THE LANDSAT THEMATIC MAPPER Given the limited success some investigators have shown with Landsat TM images of reef ecosystems (e.g., Mumby et al., 1998), it is worth integrating our modeled spectral radiances over the relevant TM bandwidths to examine how well suited the TM instrument’s radiometry is for this purpose. This type of analysis needs to be done in the context of the expected dynamic range. When designing a satellite sensor for a specific remote sensing task, one needs to know the expected radiance range from the scenes of interest so that the detector sensitivity and analog-digital converter can be chosen accordingly. For example, the Landsat TM was designed to show as much distinction as possible between various cloud-free land surface types, and the dynamic ranges of the TM1 (450–520 nm) and TM2 (520–600 nm) bands were set at approximately (prelaunch values) 0–11.1 and 0–24.7 W m⫺2 sr⫺1, respectively (Barker, 1984). Clouds usually give rise to larger backscattered radiances, but cloud mapping for weather or climate study was not part of the Landsat mission and clouds satu-
132
Lubin et al.
Figure 6. The spectral radiance at the top of the Earth’s atmosphere, in the near-nadir direction (polar angle 11.4⬚), as a function of water column depth, for A) sand, B) coralline algae, ) green macroalgae, D) algal turf, E) M. cavernosa, F) A. Palmata, G) D. stokesii, and H) S. siderea.
rate the TM1 and TM2 bands. When using Landsat for coral reef mapping, we are restricted to a sensor whose dynamic range has been optimized for general use over all tropical and temperate-latitude land surface scene types. It is worth investigating how much of the TM1 and TM2 dynamic range we are actually using when viewing coral reefs. We integrated the top-of-atmosphere spectral radiances of Figure 6 over the Landsat TM1 and TM2 bands, weighting the spectral radiance by the instrument response functions of these bands (Barker, 1984). For the eight objects considered in this study, these integrated radiances are shown as a function of water column depth for the TM1 band (Figs. 8A,B) and the TM2 band (Figs. 8C,D). The radiances are given in W m⫺2 sr⫺1 on one vertical axis, and in 8-bit counts as used by the TM digitization (scale of 0 to 255 over the full dynamic range of each sensor band) on the other vertical axis. We see that the TM1 band is already nearly ideally optimized for coral reef identification. The brightest scene of interest—sand at the shallowest depth—yields a radiance corresponding to 211 counts, and
the radiance from this scene type decreases by 95 counts (approximately 1/3 of the sensor’s dynamic range) as water column depth increases from 1 m to 20 m. The radiance from the coralline algae scene spans a range of 121–98 counts as water column depth increases from 1 m to 20 m. The radiance from algal turf spans a similar range, but is slightly less sensitive to water column depth. The four living corals, and the green macroalgae, yield radiances that encompass the lower third of the TM1 band’s dynamic range. The radiance from A. palmata actually increases slightly with water column depth. A. palmata has such a low intrinsic reflectance that, at the shallower depths, more photons are absorbed than in a semiinfinite water column. From Figures 8A,B we conclude that the Landsat TM1 band should be well suited for distinguishing living coral reefs from most surrounding objects, at least from a radiometric perspective, because the range of backscattered radiances from the objects of interest makes full use of this band’s dynamic range. Figures 8C,D show that the Landsat TM2 band is less
Spectral Signatures of Coral Reefs: Features from Space
133
Figure 6. Continued.
well optimized for coral reef identification. The radiance from sand corresponds to only 98 counts at water column depth 1 m, decreasing rapidly to 31 counts at water column depth 10 m. There is a useful contrast (tens of counts) between the other three scene types for shallower water column depths, but by depth 10 m the radiances from all eight objects occupy only the lower 15% of the TM2 band’s dynamic range. For this exercise, we extended our calculations for all eight objects to a depth of 100 m, to verify convergence of radiances over deep water. In the TM1 band, the radiances converge to a mean value of 3.052 W m⫺2 sr⫺1, with a difference of no more than 3% between bands. Most of this difference is due to the radiance over sand, which is slightly larger than that of all other objects. In the TM2 band, the radiances converge to a mean value of 1.717 W m⫺2 sr⫺1, with a difference of no more than 0.8% between bands. It is also worth investigating if the differential reflectances between the TM1 and TM2 bands contain information about object discrimination or about water column depth. First, we examine the water-leaving radiances to see if any information is present in principle; then we examine
the top-of-atmosphere radiances to see what might be realized in practice. In Table 1, we show the reflectance differences at the ocean surface between the TM1 and TM2 bands (as they might be measured by a TM “simulator” aboard a low-altitude aircraft), as a function of water column depth and for each coral reef object. Over the shallowest depths, this reflectance difference is negative, and for all objects the reflectance difference eventually becomes positive with increasing depth. At the shallowest depths, there does not appear to be much obviously useful information for distinguishing between coral species, or for distinguishing coral from coralline algae or algal turf. However, there may be useful information about water column depth in the reflectance difference over sand. The reflectance difference for sand increases noticeably with intermediate water column depth (3–7 m), toward a maximum at 10 m. If the reflectance difference over sand can be used to make a first-order estimate of water column depth (after the absolute scene reflectance in either band has been used to identify sand), then perhaps other reflectance differences could help distinguish corals from coralline algae or algal turf. The success of such an attempt would depend on the
134
Lubin et al.
Figure 7. Upwelling spectral radiance at the air-water interface, in the near-nadir direction (polar angle 11.4⬚), as a function of water column depth: A) sand, B) coralline algae, C) green macroalgae, D) algal turf, E) M. cavernosa, F) A. Palmata, G) D. stokesii, and H) S. siderea.
existence of homogeneous scene types on the scale of a TM pixel (of order 30 m). In Table 2, we show the reflectance differences as they would be measured from space. For most objects and depths, the reflectance differences between objects and depths are preserved above the atmosphere, although they are shifted positive by about 3%. The noise-equivalent reflectance in the TM instrument is of order 0.2% (Barker, 1984). Therefore, if the information in Tables 1 and 2 is useful for coral reef mapping, most of this information should be detectable by TM. DISCUSSION The radiative transfer modeling studies presented here support the results of previous case studies using satellite data (e.g., Mumby et al., 1998); previous studies have shown that sand is relatively easy to distinguish from coral reef objects, while individual reef objects are usually difficult to distinguish from one another. Due to the continuous gradation in both spectral and broadband reflectance
throughout various reef components, this result is true even for a sensor such as TM which proves to be radiometrically very well suited for imaging coral reefs. Due to water column attenuation at wavelengths longer than 600 nm and obscuration by atmospheric Rayleigh scattering at wavelengths shorter than 500 nm, the intrinsic spectral signatures of various reef components are somewhat challenging to use from the vantage point of a satellite. Radiative transfer calculations of the type presented here can illustrate the required dynamic range and sensitivity required for a satellite instrument to detect the spectral signatures of coral reef objects. Additionally, the radiative transfer calculations presented here are for idealized conditions: a maritime background tropospheric aerosol burden, and clear water column attenuation as specified by Smith and Baker (1981). For those coral reefs found in turbid waters (e.g., Miller and Cruise, 1995), the obscuring effects of higher water column attenuation might be more severe than those discussed here. If the oceanographic community were to undertake
Spectral Signatures of Coral Reefs: Features from Space
135
Figure 7. Continued.
the daunting but worthwhile task of mapping and monitoring all of the world’s reefs, radiative transfer studies of the type presented here would have to be carried out in much greater detail. In addition to higher water turbidity, other radiative effects and sources of error include varying aerosol opacity, effects of ocean surface roughness, and the effects of errors in the knowledge of mean ocean depth for a given scene. During rough water conditions, the reefs will become obscured with breaking waves and whitecaps. While this would seriously reduce our ability to extract useful biooptical data, the information might be used to “mark” or confirm the position of shallow water shoals which might help to confirm the positions of reefs in remote regions. Another fundamental design criterion concerns the orbital deployment. Commercial remote sensors such as Landsat TM or SPOT are on polar-orbiting spacecraft. Because the sensors’ flight times over tropical regions are only a small fraction of the total orbital period, a large number of orbits would have to be considered and a large number of images acquired in order to provide adequate composites for coral reef monitoring over large geographic
areas. A thorough survey of all of the world’s reefs might not be cost-effective using these commercial sensors. The recent deployment of Landsat 7 may partially remedy the cost-effectiveness problem; as part of this spacecraft’s Long Term Acquisition Plan, some well-known coral reefs are in the ongoing data collection queue and can thus be monitored effectively. However, Landsat 7’s onboard recorders cannot be programmed to collect data everywhere; there will always be a scarcity of Landsat 7 data over remote tropical regions where many coral reefs have yet to be identified, let alone monitored. A polar orbit enables an instrument to overfly the same point on the Earth’s surface every few days. This might be an advantage for monitoring specific scenes. However, because present spacecraft are Sun-synchronous, at any given low latitude location there will be only one or two potentially useful images per day, and many of them can be expected to suffer cloud contamination at the time of overflight. Orographically or thermally induced low cloud formation often occurs when the satellite is on the scene. This makes the use of commercial sensors such as Landsat
136
Lubin et al.
Figure 8. The top-of-atmosphere near-nadir radiance for all four scene types, as a function of water column depth: A) noncoral objects integrated over the Landsat TM1 band; B) corals integrated over the Landsat TM1 band; (C) noncoral objects integrated over the Landsat TM2 band; D) corals integrated over the Landsat TM2 band.
TM very costly and cumbersome for a global coral reef project, but may not be a limitation with a dedicated instrument where the image extraction and processing procedures are set up solely for coral reef studies. A low inclination orbit would enable the instrument to view the tropics throughout most of its operation, and would allow the instrument to view a given geographic region at many differ-
ent daylight hours (thus partially circumventing the problem of cloud contamination). This might be an advantage for a mission intended primarily to identify reefs and to make the first complete world map of their distribution. One possible limitation with a low inclination orbit is that the ground track would rarely repeat itself. To provide complete geographic coverage, the instrument would need
Table 1. Difference in Ocean Surface Reflectance (Scaled Radiance in Percent) between Landsat TM1 and TM2 Bands Depth (m)
Sand
Coralline Algae
Green Macroalgae
Algal Turf
Montastrea cavernosa
Acropora palmata
Dichocoenia stokesii
Siderastrea siderea
1 2 3 4 5 7 10 20
⫺1.55 2.32 4.77 6.40 7.56 8.75 9.32 7.86
⫺2.19 ⫺0.37 0.91 1.82 2.52 3.34 3.92 3.82
⫺1.03 ⫺0.60 ⫺0.27 ⫺0.01 0.21 0.51 0.81 1.21
⫺2.73 ⫺1.21 ⫺0.14 0.64 1.25 1.98 2.55 2.74
⫺3.36 ⫺1.86 ⫺0.79 0.00 0.62 1.38 2.00 2.37
⫺2.08 ⫺1.53 ⫺1.11 ⫺0.77 ⫺0.49 ⫺0.09 0.31 0.91
⫺2.43 ⫺1.57 ⫺0.93 ⫺0.46 ⫺0.06 0.43 0.88 1.36
⫺2.41 ⫺1.65 ⫺1.08 ⫺0.65 ⫺0.29 0.18 0.63 1.18
Spectral Signatures of Coral Reefs: Features from Space
137
Table 2. Difference in Top-of-Atmosphere Reflectance (Scaled Radiance in Percent) between Landsat TM1 and TM2 Bands Depth (m)
Sand
Coralline Algae
Green Macroalgae
Algal Turf
Montastrea cavernosa
Acropora palmata
Dichocoenia stokesii
Siderastrea siderea
1 2 3 4 5 7 10 20
0.93 4.34 6.35 7.62 8.41 9.26 9.53 8.20
0.82 2.45 3.51 4.23 4.75 5.31 5.65 5.42
2.53 2.88 3.16 3.37 3.59 3.80 4.04 4.36
1.08 2.33 3.21 3.86 4.40 4.96 5.43 5.58
⫺0.14 1.29 2.23 2.88 3.36 3.91 4.31 4.45
1.09 1.60 1.96 2.23 2.46 2.74 3.02 3.42
1.38 2.09 2.62 3.01 3.36 3.73 4.10 4.48
1.40 2.03 2.50 2.85 3.19 3.53 3.89 4.33
a large cross-track swath. This would most likely require the added cost and complexity of a mechanically scanning instrument, as opposed to a simpler “push-broom” device, or a more advanced CCD with a much wider field of view. At present, one excellent platform for a coral reef mission might be the International Space Station (ISS), deployed at an orbital inclination of 51.6⬚. The ISS would offer an instrument a relatively large duty cycle over tropical latitudes, and the cost of the spacecraft itself would be eliminated (except for the costs of space qualification for the instrument). The potential for decades of operation, routine maintenance by ISS crew, and even interaction with the crew for data collection, compares very favorably with a single small spacecraft whose low cost might limit the design lifetime to less than 5 years. This research was supported by the National Oceanic and Atmospheric Administration Climate and Global Change Marine Ecosystem Response Program, Award NA36GP0420. We thank the Caribbean Marine Research Laboratory for assistance with laboratory facilities and sampling. REFERENCES Barker, J., Ed. (1984), LANDSAT-4 Science Investigations Summary, Including December 1983 Workshop Results, NASA Conference Publication 2326, National Aeronautics and Space Administration, Scientific and Technical Information Branch, Washington, DC. Bour, W., and Pichon, M. (1997), Discrimination of scleractiniandominated from other reef communities using SPOT satellite imagery. In 8th Int. Coral Reef Symp. (H. A. Lessios and I. G. Macintyre, Eds.), Smithsonian Tropical Research Institute, Balboa, Panama, pp. 1487–1490. Chandrasekhar, S. (1960), Radiative Transfer. Dover, New York, 393 pp. Dustan, P., and Halas, J. (1987), Changes in the reef-coral population of Carysfort Reef, Key Largo, Florida, 1975–1982. Coral Reefs 6:91–106.
Hatziolos, M., Hooten, A. J., and Fodor, M., Eds. (1998), Coral Reefs: Challenges and Opportunities for Sustainable Management, The World Bank, Washington, DC. Holden, H., and Ledrew, E. (1999), Hyperspectral identification of coral reef features. Int. J. Remote Sens. 20(13):2545–2563. Hughes, T. P. (1994), Catastrophic phase shifts and large-scale degradation of a Carribbean coral reef. Science 265: 1547–1551. Jin, Z., and Stamnes, E. (1994), Radiative transfer in nonuniformly refracting media: atmosphere–ocean system. Appl. Opt. 33: 431–442. King, A. K. (1995), Digital analysis of a patch reef community: Rainbow Gardens Reef, Lee Stocking Island, Exuma Cays, Bahamas, M.S. thesis, College of Charleston, 92 pp. Kneiszys, F. X., Shettle, E. P., Abreu, L. W., et al. (1988), Users Guide to LOWTRAN 7, Air Force Geophysics Laboratory, Report AFGL-TR-88–0177, Hanscom AFB, MA, 137 pp. Lubin, D., and Weber, P. G. (1995), The use of cloud reflectance functions with satellite data for surface radiation budget estimation. J. Appl. Meteorol. 34:1333–1347. Lyzenga, D. R. (1981), Remote sensing of bottom reflectance and water attenuation parameters in shallow water using aircraft and Landsat data. Int. J. Remote Sens. 2:71–82. Mazel, C. H. (1997), Diver-operated instrument for in situ measurement for spectral fluorescence and reflectance of benthic marine organisms and substrates. Opt. Eng. 36:2612–2617. Maritorena, S., Morel, A., and Gentili, B. (1994), Diffuse reflectance of oceanic shallow waters: influence of water depth and bottom albedo. Limnol. Oceanogr. 39:1689–1703. Miller, R. L., and Cruise, J. F. (1995), Effects of suspended sediments on coral growth: evidence from remote sensing and hydrologic modeling. Remote Sens. Environ. 55:177–187. Mumby, P. J., Green, E. P., Clark, C. D., and Edwards, A. J. (1998). Digital analysis of multispectral airborne imagery of coral reefs. Coral Reefs 17:59–69. Smith, R. C., and Baker, K. S. (1981). Optical properties of the clearest natural waters. Appl .Opt. 20:177–184. Stamnes, K., Tsay, S.-C., Wiscombe, W., and Jayaweera, K. (1988). Numerically stable algorithm for discrete-ordinatemethod radiative transfer in multiple scattering and emitting layered media. Appl. Opt. 27:2502–2509.
