NASA Geospatial Interoperability Office Jet Propulsion Laboratory, California Institute of Technology 20040408 Release 2 remote-sensing image Pasadena, CA Jet Propulsion Laboratory, California Institute of Technology http://onearth.jpl.nasa.gov/index.html http://onearth.jpl.nasa.gov/browse.cgi?zoom=&layers=global_mosaic http://onearth.jpl.nasa.gov/browse.cgi?zoom=&layers=global_mosaic&styles=visual http://onearth.jpl.nasa.gov/browse.cgi?zoom=&layers=global_mosaic_base http://onearth.jpl.nasa.gov/browse.cgi?zoom=&layers=global_mosaic_base&styles=visual http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic&styles=&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic&styles=visual&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=visual&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=Pan&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=Blue&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=Red&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=Green&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=IR1&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=IR2&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=IR3&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=ThH&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 http://wms.jpl.nasa.gov/wms.cgi?request=GetMap&width=1000&height=500&layers=global_mosaic_base&styles=ThL&srs=EPSG:4326&format=image/jpeg&bbox=-180,-90,180,90 Multispectral mosaic of 8208, orthorectified, Landsat7 ETM+ scenes from the NASA Scientific Data Purchase GeoCover 2000 To provide Open Geospatial Consortium(r) Web Map Service(WMS) access to a medium resolution (up to 15m), multispectral and recent dataset with global extent, suitable as a base-map for many applications. To enable analysis using false-color rendering of Landsat imagery by any WMS client. False-color rendering is by the user's choice of any combination of the ETM+ bands using OpenGIS(r) Style Layer Descriptor (SLD). 19990629 20030114 date of imagery acquisition Complete Not planned -180 180 84 -60 none Landsat GCMD Parameter Keywords EARTH SCIENCE > LAND SURFACE > Topography > Landforms EARTH SCIENCE > RADIANCE OR IMAGERY > Infrared Wavelengths > Brightness Temperature EARTH SCIENCE > LAND SURFACE > Landscape > Landscape Pattern EARTH SCIENCE > RADIANCE OR IMAGERY > Infrared Wavelengths > Infrared Imagery EARTH SCIENCE > RADIANCE OR IMAGERY > Sensor Characteristics > Ultraviolet Sensor Temperature EARTH SCIENCE > LAND SURFACE > Land Use/Land Cover > Land Classes EARTH SCIENCE > LAND SURFACE > Surface Radiative Properties > Reflectance EARTH SCIENCE > RADIANCE OR IMAGERY > Visible Wavelengths > Visible Imagery ISO 19115:2003 Landsat MD_TopicCategoryCode: ImageryBaseMapsEarthCover None Global GCMD Location Keywords GREENLAND LAND SURFACE AFRICA AUSTRALIA EUROPE SOUTH AMERICA ASIA GLOBAL NORTH AMERICA none There are no restrictions as to the use of this dataset. We ask that the NASA Geospatial Interoperability Program be identified as the source if used in a publication. NASA Geospatial Interoperability Office Myra Bambacus NASA Geospatial Interoperability Program Manager mailing and physical address

Goddard Space Flight Center, Code 902

Greenbelt MD 20771 USA 301-286-3215 myra.j.bambacus@nasa.gov The WMS Global Mosaic project was conceived by George Percivall as the combination of the OpenGIS(r) Web Map Service (WMS) with the NASA GeoCover 2000 dataset. Ron Birk, NASA Earth Science Applications Division, authorized a trade-off study of several approaches which led to the selection of JPL. Lucian Plesea, JPL had previously developed a contigous US mosaic and made it accessible using OpenGeospatial WMS. The WMS Global Mosaic project was conducted under the direction of Myra Bambacus, Program Manager, NASA Geospatial Interoperability Office. The Project Manager was George Percivall, NASA Geospatial Interoperability Office (percivall@gst.com). The Principal Investigator was Lucian Plesea, California Institute of Technology, Jet Propulsion Laboratory, (Lucian.Plesea@jpl.nasa.gov). David Curkendall, California Institute of Technology, Jet Propulsion Laboratory (David.W.Curkendall@jpl.nasa.gov), acted as Senior Project Advisor none unclassified none The WMS Global Mosaic dataset was developed at JPL. In the development environment, the WMS Global Mosaic is stored as three separate datasets. One holds the six bands from the visual and near IR at 1 arc-second per pixel. The second dataset contains the panchromatic band at 0.5 arc-second per pixel. The third dataset contains the two versions of the thermal band data at 2 arc-second per pixel, for high and low gain respectively. Each dataset covers the globe from W180 to E180 and from S85 to N85, having 1,296,000x612,000, 2,592,000x1,224,000 and 648,000x306,000 pixels respectively. The image data is stored in a tiled, compressed format, with a tile size of 512x512 pixels. Each tile contains information from one band, at 8 bits per pixel. The compression is achieved by using zlib http://www.zlib.org, a freeware lossless compression library. For WMS server use, each dataset contains reduced resolution versions of the base resolution, using the same storage format. The required amount of storage for the complete mosaic, including the lower resolution copies is 1.66 TB. The panchromatic dataset is 686 GB, the multi-spectral dataset is 963 GB and the thermal is 58GB. The storage format and the associated libraries work in conjunction with the SGI Image Library tools and provide access and trans-coding to other image formats are available. JPL/Caltech retains the intellectual property rights to the software. The mosaic was constructed from the GeoCover 2000 dataset. The WMS Global Mosaic is a high level data, affected by the accuracy of the input scenes, plus the ones introduced by the mosaic building procesing step. The coordinate conversion from UTM to unprojected was done using code derived from the freeware library PROJ4 and bilinear interpolation. Average values of each of the input spectral bands and ratios between bands of a Landsat scene have been preserved as much as possible. The values of the source scenes have been altered, to reduce the visible artifacts of an image mosaic. Where two or more scenes contribute to the value of a pixel, a normalized weighted sum based on distance to scene edge was used to produce the value of the output pixel. In addition to the accuracy of the input scene set, described in the NASA document NAS13-02032, the mosaic process adds a few sources of error. The transformation from UTM to geographical coordinates is done using a tranformation derived from the PROJ4 library. The bilinear sampling method used to produce the output values also introduce a +- 1/2 pixel error at the base level of the dataset. In addition, the bounds calculation of both the mosaic process is within 1 pixel accuracy at the base level. For on-line WMS access, bilinear interpolation, bounds calculation, coordinate transformation and scaling, pan-sharpening and other spatial image processing may further affect the accuracy. Not available. The vertical accuracy of data used to geocorrect the source scenes are described in the NASA document NAS13-02032. Earthsat/NASA 2003 remote-sensing image magnetic tape Global Ortho Landsat The JPL mosaic building process is automated. In addition to the input scenes, in the original UTM projection, the code takes two control inputs. The first one is a metadata file that contains selected data about the input images, including file format, locations and statistical brightness information about each band in each input scene. The second input is a set of brightness corrections that need to be applied to each scene before they are assembled. The brightness correction needs to be generated external to the mosaic code. Internal to the mosaic process, there are three steps. First step applies the brightness correction to each band of every scene. This correction uses a first-degree polynomial, whose multiplicative and additive coefficients are applied to every pixel. Since this operation can produce overflow of the eight bit data range, the result is clipped at 0 and 255. The second step is a spatial morphing from the initial UTM projection to the geographical projection. This morphing also includes scaling to the requested output resolution. This step is required to bring the scenes into a common projection. The reprojection step uses a bilinear sampling. The UTM to geographical coordinate transformation is done using a modified PROJ4 implementation. The last step is the inter-scene blending. This step serves two functions. The first one is selecting the data pixels out of the input image space. The second function is the blending between overlapping scenes. The horizontal overlap of Landsat scenes is very small close to the equator, and increases rapidly at higher latitudes. This can produce areas where five or more scenes overlap. The weighted blending uses a data mask that marks the validity of input data. This is represented as a grayscale image that matches the input scene, with values that go between zero, (no data), to 255 (valid data). A linear variation from 0 to 255 close to the edges of a scene generates a smooth transition between one scene and the next. The value of the output pixel is computed using a normalized weighted sum of the input scenes values, using the values from the blend masks being used as the weights Global Ortho Landsat 20050525 Lucian Plesea Jet Propulsion Laboratory, California Institute of Technology Senior Staff mailing and physical address

4800 Oak Grove Dr. MS 306-463

Pasadena CA 91109 USA 818-354-3928 818-393-6141 Lucian.Plesea@jpl.nasa.gov Raster Pixel 1224000 2592000 0.0001388888888888 0.0001388888888888 Decimal degrees WGS84 WGS84 6378137.0000000 298.26 none none none The ETM+ instrument on Landsat-7 observes the Earth with 8 different filters or "bands". Bands 1, 2, 3, 4, 5, and 7 are sensitive to light energy from the sun reflected by the surface of the Earth. Each band is sensitive to a different part of the reflected solar energy. The parts of the reflected energy are defined by the length of the light waves. Thus, band 1 of the ETM+ instrument records reflected light energy only in the range of 0.45 microns (a micron is one millionth of a meter long) to 0.52 microns. The human eye sees reflected light in that band of wavelengths as the color blue; hence, band 1 is sometimes referred to as the blue band. In a similar manner, bands 2 and 3 of the ETM+ instrument records reflected green and red light, respectively. ETM+ bands 4, 5, and 7 record reflected light in wavelengths that human eyes cannot detect. These bands are referred to as near infrared (NIR, band 4) and short wave infrared (SWIR, bands 5 and 7). Band 6 of the ETM+ instrument is different from all the other bands because it does not record reflected light energy, but rather heatenergy emitted by the Earth's surface. In addition to these bands, the ETM+ instrument also has an eighth band, called the panchromatic sharpening band. ETM+ band 8 is sensitive to reflected light energy across a broad range of wavelengths that includes blue, green, red and near infrared. This panchromatic band has a spatial resolution of 15 meters, rather than the 28.5 or 30 meters of bands 1, 2, 3, 4, 5 and 7. for further information see http://landsat7.usgs.gov/ 20040408 20050525 Jet Propulsion Laboratory, California Institute of Technology Lucian Plesea Senior Staff mailing and physical address

4800 Oak Grove Drive, MS 306-463

Pasadena CA 91109 USA 818-354-3928 818-393-6141 Lucian.Plesea@jpl.nasa.gov FGDC CSDGM FGDC-STD-001-1998 Jet Propulsion Laboratory, California Institute of Technology Lucian Plesea Senior Staff mailing and physical address

4800 Oak Grove Drive, MS 126-104

Pasadena CA 91109 USA 818-354-3928 818-393-6141 Lucian.Plesea@jpl.nasa.gov WMS Global Mosaic none TIFF 6.0 geotiff The WMS Global Mosaic is partitioned into files containing data from a single band for a small spatial region. Each such file has 28800x28800 pixels at 8 bits per pixel, covering a 4x4 arc-degrees area at 0.5 arc-second per pixel for the Landsat 7 panchromatic band or 8x8 arc-degrees at 1 arc-second per pixel for the visual and infrared bands. There are 1790 panchromatic files and 3246 visual and infrared files. The geotiff files contain uncompressed data. For ease of download, the files are compressed using the GNUzip compression utility. Each file decompresses to the same size, 829670817 bytes. The transfer size advertised here is the sum in Megabytes of all the compressed file sizes. The uncompressed size for all the files in this dataset is 4TB 1360750 onearth.telascience.org http download sun4u sparc SunOS 5.9 San Diego State University, Telascience John Graham Senior Research Scientist mailing and physical address

5500 Campanile Drive, CSL 120

San Diego CA 92182 USA 619-594-5960 619-594-4372 johng@telascience.org none