Page 95
5
Conclusions and Recommendations
The National Research Council charged the Committee on NASA's Polar Geophysical Data Sets to review NASA's strategy for providing derived geophysical data sets to the polar science community. The committee was asked to provide a brief review of the strategy, scope, and quality of existing polar geophysical data sets; suggest ways to make these products and future polar data sets more useful to researchers; and consider whether the products are reaching the appropriate communities. This chapter summarizes the lessons learned in earlier chapters and provides recommendations, where possible, to guide future development of NASA and other polar data sets. The committee's conclusions and recommendations are consensus opinions based on the members' expertise and experience, and derived from committee deliberations, conversations with invited speakers (including our NASA liaison and liaisons from each of the DAACs and their user groups), analysis of the survey responses, and review of available literature.
The committee's conclusions and recommendations fall into three categories:
1. key data gaps and the highest priorities for measurements needed in the context of the polar variations of NASA's Earth Science Enterprise (ESE) science-driving questions;
2. the general NASA strategy for enhancing and providing relevant data sets to the polar science community;
Page 96
3. issues relating to the effectiveness of the Distributed Active Archive Center (DAACs) and their data distribution activities.
DATA GAPS AND MEASUREMENT NEEDS
Table 3-1 compared the science-driving questions to the types of measurements needed to address the questions and Chapter 4 then assessed available data sets and whether they are adequate to fully support the ESE. Chapter 4 then concluded with a discussion of specific measurements that would enhance NASA's present and future contributions to ESE science. In this section, the committee presents its view of the highest priorities for measurement in the cryosphere1 and data set assessment based on (1) previously identified gaps between data needs and availability; (2) the potential scientific payoffs if these gaps are filled; and (3) the feasibility and likelihood of significant progress over the next several years. The measurements identified here are generally the ones that appear repeatedly in Table 3-1, but are currently lacking or deficient in ways that present significant obstacles to progress on the fundamental ESE science issues. These 10 high-priority measurements are:
-
polar precipitation,
-
surface albedo,
-
freshwater discharge from terrestrial regions,
-
all-sky surface temperature,
-
surface turbulent fluxes,
-
permafrost,
-
ocean surface salinity,
-
ice sheet mass flux,
-
land surface characteristics, and
-
sea ice thickness.
In some cases, products that appear to be lacking (e.g., surface radiation fluxes) are under development with current sensors, so these are not included in this list. For all variables, the need is for data sets sufficient to determine the spatial and temporal variations. For those variables deemed suitable for monitoring, temporal continuity across instruments must be a key consideration.
1 The committee's deliberations and this list of 10 high priority measurements focuses directly on the cryosphere (i.e., surface characteristics) and the fluxes that determine cryospheric characteristics.
Page 97
-
Polar Precipitation—NASA should attach high priority to testing and flying appropriate sensors to measure precipitation, especially in its frozen form, in the polar regions. High-latitude precipitation data sets are needed by many segments of the research community; polar snowfall and rainfall represent a major gap in the polar geophysical database. Key needs include space-time distributions of precipitation rates for moisture budget determinations. Existing sensors (e.g., millimeter-wavelength radiometers) cannot measure precipitation over ice and snow surfaces. For uses in surface budget and hydrologic applications, fields should include the precipitation rate and the water equivalent of snow. Over ice sheets, airborne ice-penetrating radar can measure accumulation layers for information about longer timescales. It will be a challenge to identify unbiased precipitation data sets and reconcile land- and space-based measurements. Land-based station gauges suffer from well-known problems of under-measurement, particularly for solid precipitation under windy conditions. Interpolation in data-sparse regions, such as the high latitudes, also creates biases related to topography and coastal proximity.
-
Surface Albedo—Fields of surface albedo must capture the progression of melt and freeze-up over spatial and temporal scales adequate for surface energy budget evaluations and for forcing and validation of models. Meeting this objective also requires surface temperature measurements with similar time and spatial resolution. Effects of sub-grid heterogeneities must be included in grid-cell averages. Recently launched instruments that measure in the visible spectrum show promise for determining (clear sky) surface spectral albedo at appropriate resolution. Focused and careful evaluation of these products over snow and ice surfaces is required.
-
Freshwater Discharge from Terrestrial Regions—Needed quantities include lateral fluxes of freshwater to polar oceans from ice sheets and glaciers (including runoff and ice discharge) as well as surface and sub-surface runoff from non-glaciated land areas. The deterioration of surface monitoring networks adds to the urgency of exploring new satellite techniques (although it may be more economical to support the surface monitoring networks). Telemetry of gauge measurements should be feasible. Accurate discharge measurements are also required to estimate transport of land-derived materials into coastal and ocean ecosystems.
-
Surface Temperature—The wide variety of applications of surface temperature data makes all-sky measurements a high priority. Routine measurements under cloudy skies are needed to eliminate the clear-sky bias in existing satellite-based surface temperature data sets. NASA can
Page 98
take steps to enhance surface temperature data sets by supporting programs that measure in situ temperatures in high latitudes. NASA should work to develop multisensor analysis approaches that combine microwave, infrared, and radar measurements, as this approach could provide improved surface temperature products in preparation for the NPP and NPOESS missions. Calibrations accurate at the extreme conditions found in high latitudes should also be encouraged as part of efforts to generate products dealing with global temperature.
Surface Turbulent Fluxes—Sensible and latent heat fluxes are needed over all polar surfaces. Impediments to progress are the lack of all-sky fields of surface temperature, surface moisture availability, and surface winds over land and ice. Satellite-derived near-surface air temperatures and relative humidities (from NPOESS) are most likely 10 years away. While waiting for more direct measurements, NASA can support analyses using existing multisensor data sets to obtain estimates of all-sky temperatures and albedo (i.e., surface radiative fluxes) and use re-analyses to infer winds and other variables as possible.
Permafrost—Needed variables are the extent of permafrost, depth of permafrost, vertical temperature profiles and timing of thaw and freeze-up of active layer. Current technologies show some promise, and efforts could be made to exploit existing sensors (e.g., scatterometers). These sensors have been designed for other purposes, however, and their configurations will need to be optimized for permafrost. Some permafrost variables will likely require multiple-sensor products. For example, an analysis combining radar backscatter, passive microwave, and infrared measurements could significantly enhance present capabilities in permafrost mapping. NASA should support experimental multisensor analyses.
Ocean Surface Salinity—For polar regions, key requirements include high-to-moderate (10 km or finer) resolution for coastal and ice edge environments and good performance in near-freezing temperatures. Existing and proposed sensors have inadequate resolution and accuracy for polar regions; further technical development is necessary.
Ice Sheet Mass Flux (Accumulation, Ablation, and Ice Discharge)—The key needs related to ice sheet mass flux are for continued measurement and interpretation to understand the variability introduced by ice accumulation and ablation, both temporally and spatially. The NASA PARCA Program has clearly shown the utility of laser altimeter measurements for detecting changes in large ice sheets; ICESAT will make these measurements over both ice sheets and smaller ice caps on an
Page 99
operational basis. The interpretation of these measurements requires an understanding of the variability introduced by accumulation and ablation, both temporally and spatially. Airborne broadband radar measurement of internal layers tied to ice cores can provide a view of the spatial patterns of recent accumulation, while measurement or estimation of ablation rates near the ice sheet margin will require a new multifaceted approach. Interferometric studies of outlet glaciers have shown evidence of rapid change in both Greenland and Antarctica; improved Synthetic Aperture Radar (SAR) data acquisition and access, as well as airborne radar measurement of ice thickness in these drainages is required to determine changes in ice sheet mass.
Land Surface Characteristics—Particular needs to increase understanding of land surfaces include information for the monitoring of changes in wetland extent and for quantitative determinations of the severity of disturbance. Required measurements are leaf area index, water storage, canopy density, structural composition of vegetation, disturbance characteristics at resolution sufficient to capture seasonal timing of disturbance events and the interannual and decadal variation in disturbance frequency. Such data sets have been produced for other regions of the globe; particular effort is needed to produce and evaluate them for polar land areas. Effort should be made to develop pan-Arctic products spanning the past 20 years. Research is under way using existing sensors; calibration and validation is needed.
Sea Ice Thickness—Routine measurements of sea ice thickness require development of remote-sensing techniques beyond those required for mapping ice age and type. New altimeter techniques show some promise and should be encouraged, especially in view of the large recent changes in sea ice mass implied by small samples of in situ data in both hemispheres. Products must be compatible with those obtainable from aircraft, submarines, and the network of moored sonars.
NASA'S GENERAL STRATEGY ON POLAR DATA SETS
The committee strongly believes that measurements should not be made in isolation and that there is a need for across-sensor and across-variable evaluations to be an integral part of remote-sensing activities. In addition, the committee found a pervasive need for evaluations and documentation of the accuracy of existing data sets (e.g., cloud, snow, and sea ice products). These themes are common to several of the following recommendations.
Page 100
-
In several respects polar data sets have been isolated from global data sets in the overall NASA data set framework. Examples are the global satellite-derived atmospheric temperature data sets and global sea surface temperature data sets versus the TOVS vertical profiles and the POLES surface air temperatures. The committee recommends that NASA make efforts to integrate polar data sets and global data sets in the architecture of the NASA data distribution system.
-
The committee was struck by the multiplicity of data sets available for some polar variables. Cloud products, sea ice coverage, and sea ice motion are examples. The committee recommends that NASA support systematic comparisons of different versions of the same product, including calibration and validation activities, the establishment of error bars, and the merger or consolidation of data sets depicting the same variables. This activity should include the determination of needs for (1) ground-based validation sites in addition to those planned for EOS; and (2) additional field campaigns directed at validation of remote sensing products.
-
In general, satellite sensors and products have not been optimized for polar regions. Examples include those used for precipitation estimation, determination of cloud properties, and sea surface salinity. The committee recommends that NASA give higher priority to sensor optimization for polar applications.
-
For some of the variables listed in the preceding section, alternative measurement approaches may be more appropriate than reliance on satellite products. Alternative and complementary approaches could include aircraft (manned and unmanned), automated underwater vehicles, and other ground-based measurements. The committee recommends that NASA explore such alternatives from a cost-benefit perspective in the planning and design of its polar programs.
-
There is a need for enhanced interaction and feedback between modelers and data providers. This interaction can occur to some extent through data assimilation activities. There is often, however, a disconnect between modelers' needs for forcing and validation data sets and the products derived from remote sensing. The committee recommends that NASA strengthen the connection between modeling and its remote sensing strategy for the polar regions.
-
The committee recommends that NASA support research to determine optimum uses of polar remote sensing measurements in
Page 101
-
Consistency within and among data sets is a necessary prerequisite for detecting, with a known degree of certainty, the effects of climate change on high-latitude systems. Such an effort would also enhance the interaction of modelers and data providers and constitute an important framework activity for developing and testing new sensor designs. The committee recommends that NASA give high priority to fostering the development of spatially and temporally coherent, internally consistent polar geophysical data sets.
-
Finally, the continued use of satellite-derived data sets in intercomparison and synthesis activities, such as those described above, argues for the long-term archiving of satellite data sets within the NASA data distribution system. The committee considers the DAACs to be the natural vehicles for long-term archival, and we recommend that long-term archival be a DAAC function.
re-analyses so these measurements are better integrated into a global context, thereby enhancing their value to ESE and reducing their insularity. For the broadest impact, this activity should entrain, in the earliest stages, the two leading global re-analysis centers, NOAA/NCEP and ECMWF.
EFFECTIVENESS OF THE DISTRIBUTED ACTIVE ARCHIVE CENTERS
As a result of its deliberations—including consideration of the 1998 National Research Council report, Review of NASA's Distributed Active Archive Centers, conversations with DAAC managers and representatives of associated user groups, and a survey of the science community—the committee learned much about the effectiveness of the polar-oriented DAACs. Unlike the recommendations in the two preceding sections, the following recommendations pertain specifically to the polar DAACs.
-
Redundancies—The committee recommends that NASA support quantitative evaluations of possibly redundant data sets to help provide a rational basis for decisions about discontinuing certain data sets. The results of coordinated evaluations should be used by the DAACs to minimize redundancies in their holdings and to streamline their distribution responsibilities.
-
Outreach—The survey responses showed a surprising lack of unawareness of the availability of holdings of the DAACs. The committee recommends that the DAACs increase the efforts to disseminate and
Page 102
-
Links—With regard to the distribution of data products by the DAACs, the committee recommends that the DAACs install more extensive web links to other global products that contain polar data, with brief descriptions of the holdings in the sites. These “pointers” should include available information about data sources, quality, and limitations, specific to the polar regions. This information can be provided through literature references or through DAAC-initiated assessments.
-
Improvements in Tools—The committee recommends that data set providers document their individual holdings more effectively, and also provide overview documentation of the broad spectrum of data sets available at their site. The addition of browse products to data sets would help users unfamiliar with the data to judge the utility of the data set for their uses. To help overcome obstacles faced when trying to combine data sets from different sensors, data set providers should increase availability of user-friendly software tools to help with tasks such as converting from one format to another and among standard grid formats. Attention should also be given to the issue of changing technology so that archived data will remain readable with future technologies.
-
Feedback—The committee recommends that data set providers offer additional opportunity for community feedback via the creation of web bulletin boards where users may comment on their experience using data at that site. This will encourage a more coherent user community, thereby facilitating the solution of problems by users without direct intervention by the data site provider.
-
P-I Web sites—The committee recommends creation of an archive of principal investigator-generated websites containing relevant data sets or information about these data sets. This type of archive will complement the data distribution activities of the DAACs and will enhance their utility as information resources for the polar community.
-
Alaska SAR Facility—While ASF received expressions of both praise and frustration from our survey respondents, the use of ASF products has been limited by data product availability, by costs, by ease of access, and by access and distribution restrictions. There has been some recent improvement in access to SAR products. Nevertheless, we share
publicize their holdings, particularly among investigators who are not involved in large NASA programs, as well as provide overview documentation of the broad spectrum of data sets available at a site.