Track I-D-3: Land Remote Sensing - Landsat Today and Tomorrow Chairs: Hedy Rossmeissl and John Faundeen, US Geological Survey, USA Scientists in earth science research and applications and map-makers have for many years been avid users of remotely sensed Landsat data. The use of remote sensing technology, and Landsat data in particular, is extremely useful for illustrating: current conditions and temporal change for monitoring and assessing the impacts of natural disasters; aiding in the management of water, biological, energy, and mineral resources; evaluating environmental conditions; and enhancing the quality of life for citizens across the globe. The size of the image files, however, raises a variety of data management challenges. This session will focus specifically on the 30-year experience with Landsat image data and will examine four components: 1) image tasking, access, and dissemination, 2) applications and use of the imagery, 3) data archiving, and 4) the future of the Landsat program.

1. Tasking, Archiving & Dissemination of Landsat Data
Thomas J. Feehan, Canada Centre for Remote Sensing, Natural Resources Canada, Canada

The Canada Centre for Remote Sensing of Natural Resources Canada (CCRS) operates two satellite ground receiving stations, the Prince Albert Satellite Station located in Prince Albert Saskatchewan and the Gatineau Satellite Station located in Cantley, Quebec. The CCRS stations provide a North American data reception capability, acquiring data to generate knowledge and information critical to resource use decision making on local, regional, national and global scales. CCRS' primary role is to provide data related to land resources and climate change, contributing to sustainable land management in Canada.

Operating in a multi-mission environment, including LANDSAT, the CCRS stations have accumulated an archive in excess of 300 TeraBytes, dating back to 1972, when CCRS started receiving LANDSAT-1 (ERST-1) data at the Prince Albert Satellite Station. Data are made available to support near-real time applications including ice monitoring, forest fire monitoring and mapping, as well as non-real time applications such as climate change, land use and topographic mapping. LANDSAT MSS, TM and ETM+ data constitute a significant portion of the CCRS archive holdings.

In addition to Canadian Public Good data use, a spin-off benefit includes the commercial exploitation by a CCRS distributor and value-added services network.

 

2. The Work of the U.S. National Satellite Land Remote Sensing Data Archive Committee: 1998 - 2000
Joanne Irene Gabrynowicz, National Remote Sensing and Space Law Center, University of Mississippi School of Law, USA

Earth observation data have been acquired and stored since the early 1970s. One of the world's largest, and most important, repositories for land satellite data is the Earth Resources Observation Systems (EROS) Data Center (EDC). It is a data management, systems development, and research field center for the U.S. Geological Survey's (USGS) National Mapping Discipline in Sioux Falls, South Dakota, USA. It was established in the early 1970s and in 1992, the U.S. Congress established the National Satellite Land Remote Sensing Data Archive at EDC. Although data have been acquired and stored for decades, the world's remote sensing community has only recently begun to address long-term data preservation and access. One such effort was made recently by remote sensing leaders from academia, industry and government as members of a federal advisory committee from 1998 to 2000. This presentation provides a brief account of the Committee's work product.

 

3. An Overview of the Landsat Data Continuity Mission (LDCM)

Bruce K. Quirk and Darla M. Duval*, U.S. Geological Survey EROS Data Center, USA

Since 1972, the Landsat program has provided continuous observations of the Earth's land areas, giving researchers and policy makers an unprecedented vantage point for assessing global environmental changes. The analysis of this record has driven a revolution in terrestrial remote sensing over the past 30 years. Landsat 7, which was successfully launched in 1999, returned operation of the Landsat program to the U.S. Government. Plans have been made for the follow-on to Landsat 7, the Landsat Data Continuity Mission (LDCM), which has a planned launch date of late 2006.

The scientific need for Landsat-type observations has not diminished through time. Changes in global land cover have profound implications for the global carbon cycle, climate, and functioning of ecosystems. Furthermore, these changes must be monitored continually in order to link them to natural and socioeconomic drivers. Landsat observations play a key role, because they occupy that unique part of the spatial-temporal domain that allows human-induced changes to be separated from natural changes. Coarse-resolution sensors, such as the Moderate-Resolution Imaging Spectroradiometer (MODIS) and the Advanced Very High Resolution Radiometer (AVHRR) are ideal for monitoring the daily and weekly changes in global biophysical conditions but lack the resolution to accurately measure the amount and origin of land cover change. High-resolution commercial systems, while valuable for validation, cannot acquire sufficient global data to meet scientific monitoring needs. Landsat-type observations fill this unique niche.

A joint effort between NASA, the U.S. Geological Survey (USGS), and the private sector, LDCM will continue the Landsat legacy by incorporating enhancements that reduce system cost and improve data quality. Following the 1992 Land Remote Sensing Policy Act, the LDCM seeks a commercially owned and operated system selected through a competitive procurement. Unlike earlier Landsat commercialization efforts, however, the LDCM procurement is based on a rigorous Science Data Specification and Data Policy, which seeks to guarantee the quantity and quality of the data while preserving reasonable cost and unrestricted data rights for end users. Thus the LDCM represents a unique opportunity for NASA and the USGS to provide science data in partnership with private industry and to reduce cost and risk to both parties, while creating an environment to expand the commercial remote sensing market.

The data specification requires the provision of 250 scenes per day, globally distributed, with modest improvements in radiometric signal-to-noise (SNR) and dynamic range. Two additional bands have been added: an "ultra-blue" band centered at 443 nm for coastal and aerosol studies, and a band at either 1,375 or 1,880 nm for cirrus cloud detection. No thermal bands will be included on this mission. Additional details are available on the LDCM specification, mission concept, and status.

* Raytheon. Work performed under U.S. Geological Survey contract 1434-CR-97-CN-40274.

4. Current Applications of Landsat 7 Data in Texas
Gordon L. Wells, Center for Space Research, The University of Texas at Austin, USA

The rapid delivery of timely information useful to decision makers is one of the primary goals of the data production and application programs developed by the Mid-American Geospatial Information Center (MAGIC) located at the University of Texas at Austin's Center for Space Research. In a state the size and nature of Texas, geospatial information collected by remote sensing satellites can assist a broad range of operational activities within federal, state, regional and local government departments. In the field of emergency management, the state refreshes its imagery basemap using Landsat 7 data on a seasonal basis to capture the locations of recent additions to street and road networks and new structures that might be vulnerable to wildfires or flashfloods. Accurately geolocated satellite imagery can be incorporated into the geographic information system used by the Governor's Division of Emergency Management much more rapidly than updated records received from the department of transportation or local entities. For many activities involving the protection and enhancement of natural resources, Landsat 7 data offer the most economic and effective means to address problems that affect large areas. Invasive species detection and eradication is a current concern of the Texas Department of Agriculture, Texas Soil and Water Conservation Board and the Upper Colorado River Authority. Invasive saltcedar is one noxious species that can be identified and removed with the help of satellite remote sensing. The information required by policy makers may extend beyond state borders into regions where satellite reconnaissance is the only practical tool available. For international negotiations involving the shared water resources of Texas and Mexico, satellite imagery has made a valuable contribution to the monitoring of irrigation activities and the local effects of drought conditions. In the future, there will be increasing concentration on shortening the time lag between the collection of instrument data by MAGIC's satellite receiving station and final product delivery in the projection, datum and file format required for immediate inclusion into operational analyses by the various agencies in the region.

5. Development of Land Cover Database of East Asia
Wang Zhengxing, Zhao Bingru, Liu Chuang, Global Change Information and Research Center, Institute of Geography and Natural Resource Research, Chinese Academy of Sciences, China

Land cover plays a major role in a wide range of fields from global change to regional sustainable development. Although land cover has dramatically changed over the last few centuries, util now there has been no consistent way of quantifying the changes globally (Nemani, and Running, 1995). Land cover dataset currently used for parameterization of global climate models are typically derived from a range of preexisting maps and atlases (Olson and Watts, 1982; Matthews, 1983; Wilson and Henderson-Sellers, 1985), this approach has several limitations (A. Strahler and J. Townshend, 1996). Another important data source is statistical report, but some statistical land cover data seems unreliable. At present, the only practical way to develop land cover dataset consistently, continuously, and at globally is satellite remote sensing. This is also true for the development of land cover dataset of East Asia.

The 17-class IGBP land cover unit includes eleven classes of natural vegetation, three classes of developed and mosaic lands, and three classes of non-vegetated lands. This system may be useful at global level, but there is a very serious shortcoming: only one class for arable land. Since the arable land is the most dynamic and important area of the man-nature system, it is essential to characterize arable land sub-system to more details.

There are still some potentials for finer classification in current 1-km AVHRR-NDVI data sets. A decision tree classifier is used to transfer all input data into various pre-defined classes. The key to accurate interpretation is to identify more reliable links (decision rules) between input data and output classes. The basic theory under the decision tree is that any land cover class should be an identical point determined by a multi-dimensional spaces, including multi temporal NDVI, phenology, ecological region, DEM, census data etc. The preliminary research shows that stratification using ecological region and DEM can simplify the decision tree structure and yield more meaningful classes in China's major agricultural regions. Arable land cover may be classified at two levels, first level describes how many times the crops are planted, and second level the crop characteristics.

The current land cover classification based on 1-km AVHRR-NDVI data sets still have serious limitations for parameterization of some models. The nominal 1-km spatial resolution images may produce quite a lot mixed pixels, but some models need pure pixel, e.g. DNDC model. However, the coming 250-m MODIS-EVI data set will narrow the gap between model need and data supply to some extent. Using the approaches developed from AVHRR, MODIS will yield more reliable land cover data of East Asia.