The L-1 orbit of Triana improves the view of the high latitudes during the sunlit part of the year (see Figures 5 and 6). This is much improved over the view of high latitude locations available from standard geostationary satellites (GOES, GMS) that are also capable of viewing from sunrise to sunset. Instruments aboard GOES or GMS have their fields of view centered on the equator, and their images of high latitudes therefore contain too much geometric distortion for many remote sensing applications. Triana¹s orbit and good spatial resolution give EPIC an ability to make major contributions to problems in Arctic atmospheric science and climate study, including stratospheric ozone depletion and UV radiation, tropospheric aerosols (the Arctic "haze"), and polar meteorology.
Arctic ozone depletion events, significant examples of which have occurred during half of the 1990s' northern-hemisphere springs (e.g., Müller et al., 1997), are more complex and geographically less extensive than the similar depletion in the Antarctic. The conventional understanding of ozone depletion in the Arctic suggests that springtime ozone depletion is not as severe as in the Antarctic due to a less pronounced northern hemisphere polar vortex (Solomon, 1999). In the northern hemisphere, greater atmospheric wave activity induced orographically by land results in a warmer stratosphere with less PSC (Polar Stratospheric Cloud) formation during winter, and earlier springtime stratospheric warmings. Dynamical considerations that have so far limited the size of Arctic ozone depletion events also render them more geographically variable. The coarse spatial resolution of TOMS is often inadequate to resolve the spatial structure of the Arctic polar vortex boundary and to follow the complete time history of an Arctic ozone depletion event that might cover a limited geographical area.
The Arctic is host to a considerable human population and more extensive land ecosystems compared to Antarctica. These may be adversely impacted by enhanced UV radiation under stratospheric ozone depletion events. Monitoring springtime Arctic ozone depletion, including the mapping of enhanced UV-B radiation at the ground (Lubin et al., 1998), will be a valuable activity for both ecological studies and public health awareness in northern high latitude communities (e.g., Scandinavia; Moan and Dahlback, 1993). The Triana 14-km resolution (at high latitudes) will provide major improvements to satellite UV monitoring capability: first in the ability to validate more accurately such remote sensing retrievals with ground-based spectroradiometer measurements made by NSF Office of Polar Programs or ARM at Barrow, Alaska and by European researchers at Tromso; second in the ability to resolve more accurately the spatial variability in the surface UV radiation field.
There is a well-known coupling between CO2-induced tropospheric warming and stratospheric cooling (Fels et al., 1980; Shindell et al., 1998). Austin et al. (1992) have shown how an increasing tropospheric CO2 burden may eventually lead to Arctic ozone depletion events that approach the severity of those in the Antarctic. EPIC will be able to monitor stratospheric ozone concentrations throughout the Arctic with high spatial resolution for several years. The observed ozone variability can be correlated with stratospheric and tropospheric temperatures and dynamics (from NCEP or ECMWF reanalyses, or infrared sounder data) and also with any observed trends in Arctic tropospheric mean temperature. This capability will enhance our understanding of the relationship between stratospheric ozone depletion and the various factors that govern temperature in the lower stratosphere.
There is a well-known anthropogenic aerosol burden in the Arctic troposphere, known as the "Arctic haze". It is now recognized that tropospheric aerosols play an important role in regional climate forcing (Kiehl and Briegleb, 1993). In the Arctic, tropospheric aerosols have an opacity that is sufficient to affect directly shortwave radiative fluxes and tropospheric heating rates (Pilewskie and Valero, 1993; Tsay et al., 1989; Valero et al., 1984, 1988). There is also a potential "indirect" radiative effect of aerosols, in which the presence of aerosols (acting as condensation nuclei) biases the cloud droplet size distribution toward a smaller effective radius, which thereby increases cloud opacity and albedo for a given liquid water path (e.g., Platnick and Twomey, 1994). This indirect radiative effect has not yet been verified by experiment in the Arctic (it is difficult to quantify empirically with standard field methods), but it must be a focus of future Arctic climate studies. The surface radiation budget is known to be sensitive to the particle size distribution in the extensive stratiform cloud cover that is a prominent feature of Arctic meteorology (Curry and Ebert, 1992).
To date, most of our knowledge about the geographic and temporal variability in Arctic haze has come from a handful of ground stations, particularly in Alaska (e.g., Shaw, 1982; Polissar et al., 1998). Alaska has proven to be a useful location for these studies, because Alaska is affected alternately by Arctic and Pacific air mass systems. In the Pacific air mass system, aerosol chemical composition is characterized by enrichment in elements related to sea salt. In the Arctic air mass system, pollutants such as excess sulfates are transported over long distances to Alaska. The abundance of these excess sulfates has been shown to exhibit a general negative gradient from northwestern to southeastern Alaska, indicating a long-distance source to the northwest of Alaska (e.g., industrial activity in or near the Russian Arctic). The overall seasonal cycle in the Arctic haze involves a maximum tropospheric aerosol burden during late winter and spring, with a decrease toward a minimum during mid-summer due primarily to removal by increased precipitation. For future climate study, it will be important to (a) better characterize the temporal and spatial variability in aerosol opacity with a better resolution than is available from a handful of ground stations, and (b) identify possible transport pathways. EPIC's unique UV/visible wavelength capability for aerosol opacity retrieval should facilitate greater understanding of the mesoscale and large-scale behavior of the Arctic haze. During the spring and summer, much of the Arctic exhibits low enough surface albedo (e.g., tundra, open ocean, broken sea-ice cover) that EPIC¹s aerosol retrieval algorithms will be effective. The high time resolution of the EPIC imagery, combined with large-scale views of the Arctic, will maximize the number of cloud-free scenes from which we can map aerosol opacity.
The Arctic is expected to exhibit a particularly sensitive response to radiative forcing from anthropogenic greenhouse gases, due to climatological mean surface and lower tropospheric temperatures near the triple point of water. Satellite passive microwave observations of Arctic sea-ice have already revealed downward trends in total sea-ice extent that are consistent with a "global warming" scenario (Cavalieri et al., 1997). Although the fundamental "ice-albedo" and "cloud-radiation" feedback mechanisms have been identified (Curry and Webster, 1999), GCM simulations of present-day Arctic climate remain inadequate. Current GCMs tend to make large errors in simulating cloud amount (Chen et al., 1995; Curry et al., 1996), and also tend to underestimate natural climate variability in the Arctic (Battisti et al., 1997).
While modern field studies such as the year-long Surface Heat Budget of the Arctic experiment (SHEBA, led by NSF Office of Polar Programs during 1997-98) are providing many important advances in our understanding of local thermodynamics, cloud microphysics, and radiation, future work must involve tropospheric dynamics. The Arctic surface radiation budget is strongly modulated by the pervasive stratiform cloud cover, and this cloud cover is governed to a large extent by the advection of warm air and moisture from lower latitudes (Curry and Herman, 1985; Pinto, 1998). The mean poleward transport of water vapor is found to be positive at most low-to-mid tropospheric levels (Serreze et al., 1995), but with considerable geographic variability, and with an understanding of interannual variability still yet to be realized. In order to fully understand the response of the Arctic climate system to possible anthropogenic changes in "greenhouse" gas forcing, it is necessary to understand the dynamical factors that govern meridional energy transport between lower latitudes and the Arctic, such that we can simulate this energy transport with GCMs. High-time-resolution imagery of large- scale cloud fields, and retrievals of water vapor, can help us reach this goal. According to Stone (1997), "A basin-wide assessment of the temporal and spatial relationship between temperature and cloud distributions is needed to verify simulations of Arctic climate. This must include an evaluation of the advective processes that impact those distributions. Because it is impossible to collect the necessary data to accomplish this task at the surface, we must rely on satellite data ultimately to make these assessments and to monitor Arctic climate in the future."
Clearly then, a satellite instrument that can provide useful imagery for synoptic meteorology in the Arctic can make important contributions. Standard geostationary instruments cannot image high latitudes without considerable geometric distortion. Existing polar-orbiting instruments, such as the Advanced Very High Resolution Radiometers (AVHRR) aboard the NOAA spacecraft, offer the potential for many images over the course of a day due to the convergence of the orbital subtracks at high latitudes. However, there is at present no unified data collection strategy in the Arctic for these spacecraft, which typically provide only line-of-sight telemetry and therefore require antennas located in the Arctic to collect Arctic data. Piecing together AVHRR (or similar) images for long time periods, with high temporal resolution (several images per day), from disparate viewing angles, and covering large geographic areas (i.e., the entire sub-Arctic and Arctic), is cumbersome to the point of near-impossibility.
During the sunlit half of the year, Triana¹s whole-Earth view overcomes these limitations. We should mention that EPIC will not be perfect for this purpose, due to its lack of thermal infrared channels. EPIC¹s limitation to wavelengths shorter than 1.1 microns will make cloud detection problematic over regions containing near-100% sea- ice cover. Over Arctic land masses during summer (e.g., tundra), and over the open ocean, clouds will be easily identified by radiance contrast with the underlying surface. Over the Arctic Ocean in the marginal ice zone (sea-ice concentrations less than 50- 60%), image texture can be used to distinguish clouds from the underlying high-albedo surface (e.g., Ebert, 1987; Lubin and Morrow, 1998). Over uniform high albedo surfaces, such as the majority of multiyear ice in the central Arctic Ocean, neither radiance contrast nor texture is entirely reliable for identifying clouds. However, for climate study, this limitation of EPIC is partially offset by the fact that we are interested mainly in tracking air masses moving from lower latitudes into the Arctic. Also, for study of Arctic ozone depletion describe above, retrieval of total column ozone abundance is not hampered by a high albedo surface. Thus, we are not claiming that Triana will be an ideal platform for all meteorological applications over the Arctic. Nonetheless, Triana will overcome many of the viewing limitations with existing satellite platforms, and in conjunction with continuous detailed surface data from the ARM site at Barrow, Alaska (Stamnes et al., 1999), should further our understanding of Arctic meteorology and climate.