1. Introduction to DMSP
2. GIS Applications of DMSP Imagery
4. Associated Links
INTRODUCTION TO DMSP
The Defense Meteorological Satellite Program (DMSP) is a series of weather satellites developed to support Department of Defense (DoD) operations. DMSP satellites are in a near-polar sun-synchronous orbit at an altitude of approximately 835 kilometers. The orbital period is 101.4 minutes which produces 14.2 orbits per day. Each DMSP satellite monitors the atmospheric, oceanographic, and solar-geophysical environment of the earth.
The Operational Line Scan (OLS) is the primary image acquisition system on DMSP spacecraft. Two spectral channels are used in the OLS. One channel responds to reflected solar or lunar radiation in the 0.4-1.1 micron range, chosen so as to provide maximum contrast between earth, sea, and cloud elements of the image field, and is termed the visible or L-band. Two different detectors are used for the collection of L-band data; one collects fine resolution data (0.3 nautical miles) and the other collects smooth resolution data (1.5 nautical miles). The other channel of the OLS responds to emitted earth, atmosphere, and cloud radiation in the 10.8-12.5 micron range (infrared). Data collection at this channel is accomplished in the same manner as for L-band data. The image data is collected at various sites and sent to Air Force Global Weather Center (AFGWC) and the National Geophysical Data Center (NGDC). At AFGWC, a polar-stereographic map is projected onto the image and the data is used in numerous ways. The IR data is 8-bit while the visible data is 6-bit (64 shades). Because of recorder limitations aboard the satellites, fine-resolution data is only available for pre-selected areas.
There are many additional environmental sensors on board DMSP satellites but the main sensor that that will be discussed here is the Special Sensor Microwave Imager (SSM/I). The SSM/I is a seven channel, four frequency passive radiometer which detects energy emitted by the earth's atmosphere in the microwave portion of the spectrum. The frequencies utilized by the SSMI (selected to achieve specific objectives for measuring parameters) are 19.35, 22.235, 37, and 85.5 GHz. The data is downlinked to AFGWC, written to 8mm tapes, and sent to NGDC for processing. At NGDC, the SSM/I pixels are geolocated using the satellite ephemeris and satellite altitude corrections. The resultant imagery is 8-bit at a resolution between 12.5 and 25 kilometers.
GIS APPLICATIONS OF DMSP IMAGERY
DMSP imagery has many applications aside from the purely meteorological aspect (cloud detection). These applications range from biomass burning detection (forest fires included here) to ice and snow detection.
City Lights: The OLS visible channel has the capability to detect the lights of cities at night, allowing for a different perspective on the spatial analysis of the urban landscape. Below is a sample of this type of application. It is interesting to note that the lights of fishing vessels were also detected in the Sea of Japan. Click here to view more images of this application.
Lightning Detection: The OLS visible channel can also detect lightning in thunderstorm clouds. Lightning is indicated by horizontal white lines in the imagery. Click here to see an example of this application.
Biomass Burning: The OLS sensor can be used to detect forest fires as well as biomass burnings. The IR channel can be utilized when clouds obscure the area of interest. Not only can the core area of a fire be detected, but the smoke plume can be seen on the imagery as well, giving the analyst an idea of which direction the smoke will travel and who could be affected by the residual smoke. Click here to view an image used to detect and estimate coverages of forest fires that occurred in Idaho during August of 1994. Click here to view images of various biomass burning events.
Snow and Ice Detection: The OLS sensor can detect snow and ice fields using both the visible and infrared channels, although analysis is more difficult in the IR mode. Snow is easily discernible from clouds - it forms a dendritic pattern in mountains and, over flat terrain, rivers and heat islands are visible. Ice is a little more difficult to discern - best seen over clear areas or when the ice has large cracks or clearly defined edges. This application can help in the determination of snow and ice accumulations and its importance in the earth system. Click here to see an example of snow in the Rocky Mountains, or here to see ice in Hudson bay and Lake Winnipeg.
(Image of the SSM/I sensor)
The SSM/I sensor has many different types of applications, from atmospheric phenomena to soil moisture analysis. Most GIS type applications of SSM/I are accomplished at the 19 GHz frequency, which is sensitive to the characteristics of the earth's surface. Due to the resolution of the sensor, these applications are of a large-scale magnitude.
Bare Soil: The SSM/I can detect and classify bare soil, based on reradiated heat from the earth's surface. Adding water to dry soil dramatically changes its radiative characteristics and appearance in SSM/I imagery. Dampened soil will have lower brightness temperatures than the same soil in a dry condition. SSM/I can also estimate soil moisture, although this is subject to a number of difficulties, ranging from the presence of vegetation to the roughness of the soil. The image below shows a large portion of the Middle East during winter in 4 SSM/I channels (clockwise from upper left are 19, 22, 85, and 37 GHz). Note how well the waters of the Persian Gulf contrast with the warmer land surrounding it except at 85GHz (85GHZ is used mainly for cloud detection). The Tigris/Euphrates basin is also much cooler than the adjoining desert, due to its increased soil moisture content. The effect of different terrain elevations and soil types can be seen as well. The linear feature in central Saudi Arabia is an elevated line of hills, and is cooler because of both its altitude and soil condition.
Vegetation: The appearance of vegetation in SSM/I imagery is a combination of modified radiation from the underlying soil and emissions from the vegetation canopy itself. Below is an example of this type of SSM/I imagery. The bright green area is lush rainforest in Brazil, Bolivia, and elsewhere in the Amazon drainage basin, which appears very warm to the SSM/I.
Snow: The SSM/I can detect snow cover at frequencies above 20 GHz. It can also estimate the depth of snow cover from 0 mm to 400 mm, although not always accurate under all snow conditions.
Sea Ice and Ice Concentration: SSM/I can detect and distinguish sea ice from its surrounding waters. Ice has a higher emissivity than water, and thus to the SSM/I looks significantly warmer than the surrounding water. Using an Ice Concentration algorithm, the fraction of an ocean area that is covered by ice can also be calculated from SSM/I data. This value is given in percent, ranging from 0% to 100% with a quantization of 5%. Below is an example of the end product of the SSM/I snow and ice function.
Surface Type: SSM/I data can be used to determine surface type based on temperature. This complex algorithm can classify surface types from glacial regions to desert regions. Below is an example of this application in southeastern Canada. Some of the categories are: FL - Flooded Soil, WG - Wet Ground, DS - Dry Snow, RG - Rain over Ground, etc.
Estimation of Ocean Surface Wind Speeds: As the wind blows across open water, it roughens the surface and produces foam. Both of these effects increase the emmisivity of the sea surface. Sea foam, in paricular, has a very high emmisivity. The estimation of ocean surface wind speeds is a complex algorithm that is normally calculated from a combination of SSM/I channels in meters per second. It should be noted that other environmental phenomena can degrade this estimation. Rain impaction on the sea surface roughens the surface and causes higher brightness temperatures, which can result in erroneous wind speeds. Also, large amounts of cloud water can alter the wind signal. Below is an example of SSM/I derived ocean wind speeds.
There are many new developments and techniques in the exploitation of DMSP data. Air Force Global Weather Center (AFGWC) has developed multispectral imagery from OLS DMSP fine data. The imagery is a colorized combination of the visible and infrared channels. the visible channel is colored yellow and the infrared channel is colored blue and then the two images are merged together. Low clouds and snow which are bright in the visible channel but not in the infrared channel result in yellow shades. High cirrus which is bright in the infrared but not in the visible result in blue shades. This multispectral imagery can be very beneficial. Relative cloud heights and depths can be determined at a glance. Non-precipitating clouds can easily be distinguished from more substantial cloudiness. When placed in a loop, low-level and high-level winds can be deduced, aiding meteorological predictions over regions that may not have conventional weather observations.
AFGWC also employs a solar elevation correction to some of its DMSP data. Visible imagery appears darker in the morning and evening than they do at noon. By calculating the geographic position of each point in the image and knowing the time the image was captured, they can compute how high in the sky the sun was. This algorithm then brightens the pixel based on that sun angle.
DMSP OLS in the visible channel can also detect the aurora, also referred to as the northern lights. This phenomena, believed to be of electrical origin, is best seen in arctic regions. To see an example of this imagery, click here.
NATIONAL GEOPHYSICAL DATA CENTER: Manages environmental data in the fields of marine geology and geophysics, paleoclimatology, solar-terrestial physics, solid earth geophysics, and glaciology (snow and ice).
NATIONAL SNOW AND ICE CENTER (NSIDC): Established by NOAA in 1982 to serve as a national information and referral center in support of glaciological reasearch. NSIDC archives digital and analog snow and ice data, maintaining information about snow cover and avalanches, glaciers and ice sheets, floating ice, ground ice and permafrost, atmospheric ice, paleoglaciology and ice cores.
DMSP SYSTEM PROGRAM OFFICE (SPO): The SPO is the Air Force agent responsible for the acquisition and sustainment of the multi-service DSMP system of polar-orbiting weather satellites.
LT. PAUL McCRONE'S (USN) SSM/I INFORMATION PAGE: An informational page on SSM/I from an AFGWC meterological satellite (METSAT) expert.
NCDC SSM/I PAGE: An excellent SSM/I information page compiled by the National Climatic Data Center (NCDC).
AFGWC METEOROLOGICAL SATELLITE PAGE: Air Force Global Weather Center's satellite page.
SSM/I SAMPLES: A good SSM/I sample site.
DOWNLOAD an mpeg of one equator to equator orbit of a DMSP satellite (F12).
REAL-TIME SSM/I WIND DATA: Real-Time SSM/I data of ocean surface wind speeds.
SEA ICE CONCENTRATIONS: A good SSM/I derived data page on sea ice concentrations.
(Written by Chris Comte)
SSM/I Baresoil, Vegetation, and Surface Type samples taken from HUGHES Systems SSM/I users guide (October 1996)
All other images were found on the internet