1.0 Overview

Is there an image for my study area?, what is the resolution?, how much does it cost?. We have all heard these questions. To answer these questions is not as straightforward as one thinks. There are numerous satellites, various data providers/sources, and there are numerous considerations like costs, copyrights, and quality. So a systematic approach to data search and acquisition is the first step to using satellite sensor data in your project.

In order to facilitate this, the IWMI?s RS/GIS lab. Prepared this guide book. This is by no means a comprehensive document to search for images. There are numerous other satellite and sensors and possibilities. However, this is definitely a document that will meet the need of most satellite sensor data users.

2.0 Where do you begin ?

As a researcher/user (simply referred here onwards as user) your first task is to get a precise bounding (e.g., upper left and lower right latitude and longitude) coordinates of your study area. Then, first and foremost check IWMIDSP (http://www.iwmidsp.org) for available images. If not, most satellites have boxes (path/row) drawn on map that you can just show your area (look for such a path/row map in RS/GIS Lab.). However, if you are getting very high resolution images you better have precise coordinates of your study area, since you will have to pay per square kilometers.

3.0 How do you look for image/s you want ?

First ask your self what is your main objective. Based on this look for ?ideal? image/s that helps in your study objectives. Decide to purchase images based on what is feasible financially. But the good news is that there is now enormous amount of free satellite images for any part of the world.

4.0 What types of satellite images suite your interest ?

All satellite images can be roughly categorized into following categories (and you can choose them based on factors such as your project objectives, resources available for processing, expertise available for interpretation, and costs). We almost exclusively consider optical (or passive) sensors in section 4.1 to 4.3. In section 5.0 we will touch upon hyper spectral sensors and in section 6.0 we will discuss active sensors. But your first gateway for all remote sensing data should be IWMIDSP (http://www.iwmidsp.org).

4.0.1 Your first gateway for satellite images!

First and foremost searchhttp://www.iwmidsp.org for almost any type of imagery (section 4.1 through 6.0). If IWMIDSP has already acquired your images, that will save you lots of time. Probably, IWMI RS/GIS unit has also processed and brought your data to a refined level that is ready to use in your project, saving you considerable amount of time and resources.

4.1 Coarse resolution (100 m and up)

There is a treasure of high quality science data from coarse resolution sensors. These are ?spatially? coarse (pixel size 100 m, 250m, 500m, 1000 m, and 10000 m), but are rich in spectral, temporal, and radiometric resolutions. Here are the sources for your coarse resolution data:

4.1.1 AVHRR pathfinder

In www.iwmidsp.org the AVHRR 10-km data is available for the entire globe- stitched as a continuous time series.

Coverage: Global;

Temporal: monthly from 1981 July through 2001 September;

Spatial resolution: 10-km or 0.1 degree;

Spectral resolution: 4 bands-1 red, 1 NIR, and 2 thermal;

Radiometric: Reflectance (0 to 100 %);

NDVI: -1 to + 1;

Download site:http://www.iwmidsp.org

Format: ERMapper .ers

Costs: free

Note: there is also a 10-day data for the same period, but not as a single mega file.

4.1.2 SPOT Vegetation

SPOT Vegetation data is free but will require a user name and password. There is usually a 3 month time lag for image release.

Coverage: Global

Temporal: 10-day and monthly composites; 1998-present

Spatial resolution: 1-km

Spectral resolution: 4 bands (1 blue, 1 red, 1 NIR, and 1 SWIR)

Radiometric: 8-bit

NDVI: -1 to + 1

Download site: http://free.vgt.vito.be/

Format: .dat binary

Costs: free

4.1.3 MODIS Terra and Aqua

MODIS daily data are available daily, but the most recommended products are 8-day and 16-day composites. Currently these are most widely used Global satellite sensor data products. Rich science quality datasets and for zero costs. This is a narrow band sensor-meaning the band width in electromagnetic spectrum are narrow compared to broad band widths of sensors like Landsat.

Coverage: Global

Temporal: 8-day, 16-day; March 2000-present

Spatial resolution: 250 to 1000-m; 500-m is most widely used and recommended

Spectral resolution: 7-land bands for 500-m (blue, green, red, 2 NIR, and 2 SWIR), 2 bands for 250-m (red and NIR); there is also original 36 band data that includes thermal bands.

Radiometric: 12-bit

NDVI: -1 to + 1

Download site:http://edcimswww.cr.usgs.gov/pub/imswelcome/

Format: .hdf

Costs: free

4.2 Medium resolution (20-100 m)

A large volume of medium resolution data is free of cost. But for some, we need to pay. Here are some details:

4.2.1 Landsat data of 1970s, 1990s, and 2000s: Historical medium resolution data:

Start here if you need free historical data of high quality. These are produced under pathfinder datasets program. These Landsat data ate orthorectified and can be used for georectification of maps and images. + 50 m accuracy.

Coverage: Global

Temporal: one time each for mid-1970s, late 1980-early 1990s, and late 2000s-early 2001s

Spatial resolution: MSS sensor for 1970s-56mx89m, TM sensor for 1990s-28.5m, ETM+ sensor for 2000s-28.5 m, with Pan 15 m

Spectral resolution: MSS 1970s: 4 bands (green, red, 2-NIR), TM 1990s: 7 bands (blue, green, red, NIR, 2 MIR, and 1 TIR), and ETM+ 2000s: 8 bands (blue, green, red, NIR, 2 MIR, and 1 TIR, and 1 Pan)

Radiometric: 8-bit

Download site:

http://glcf.umiacs.umd.edu/index.shtml

Format: GEOTIF

Costs: free

4.2.2 Landsat Geocover 1990s, and 2000s: Historical medium resolution data

Start here if you need base maps/images. The Geocover products are just 3 band data. If you want 7 band data go to section 4.2.1 above. But the data you need is for planning/reconnaissance/base maps then use Geocover products (less data volume). These are orthorectified and hence are geographically ?perfect? to + 50 m accuracy.

Coverage: Global

Temporal: one time each for 1990s and 2000s

Spatial resolution: TM sensor for 1990s-28.5m, ETM+ sensor for 2000s-28.5 m, with Pan 15 m

Spectral resolution: TM 1990s: 3 bands (SWIR2, NIR, Green), and ETM+ 2000s: 3 bands (SWIR2, NIR, and Green)

Radiometric: 8-bit

Download site:

http://glcf.umiacs.umd.edu/index.shtml

Format: MrSid

Costs: free

4.2.3 Landsat data of recent times and historical: Historical medium resolution data

This is a potential source of cheap Landsat data for new acquisitions. When someone buys a new image they can share the image in Landsat.org so that others can get it for $50 (do some public service)!. But not all available data is available here (for that go to point 4 below).

Coverage: Global

Temporal: 8 to 16-days

Spatial resolution: 28.5 m multi spectral, with Pan 15 m

Spectral resolution: 8 bands for 2000s (blue, green, red, NIR, 2 MIR, and 1 TIR, and 1 Pan)

Radiometric: 8-bit

Download site:

http://landsat.org/

Format: GEOTIF

Costs: $50 to $600 per scene

4.2.4 Landsat data of recent times and historical from USGS: Historical medium resolution data

This is actually the real official source of Landsat data. But section 4.1 to 4.3 gives option for cheaper Landsat data. Look there first and if you do not find any Landsat images from section 4.1 to 4.3, visit this.

Spectral resolution: 8 bands for 2000s (blue, green, red, NIR, 2 MIR, and 1 TIR, and 1 Pan)

4.2.5 Landsat (recent and historical) data from commercial sources of Spaceimaging: Medium resolution data

This is potentially the costlier place to purchase Landsat images. But if you do not find images of interest in section 4.1 to 4.4, this may be the place to search.

Spectral resolution: 8 bands for 2000s (blue, green, red, NIR, 2 MIR, and 1 TIR, and 1 Pan)

4.2.6 SPOT HRV data: Historical and recent

SPOT has an excellent service and is probably the best place to go if you plan to program a satellite for new acquisitions. You are most likely to have successful acquisition for specific sites and for specific time periods.

Spectral resolution: 4 multispectral (blue, green, red, NIR, and at times green, red, NIR, and MIR), 1 Pan band, and stereoscopic options

Note: This is actually the real official source of Landsat data. But point 1 to 3 gives option for cheaper Landsat data.

4.2.7 ASTER data

ASTER sensor is also on the same satellite as MODIS. Excellent data for cheap prices. Availability not easy. But worth looking at. Attractive to get stereo pairs (for DEM generation) for a nominal prize.

Spectral resolution: 15 bands multispectral (2 visible, 2 NIR, 6 SWIR, and 5 TIR)

4.2.8 ALI data

ALI is next generation Landsat data. Site you must visit. Quality of data much better than Landsat. High quality cheap data.

4.2.9 IRS LISS data: Historical and recent

Take a look at what is available in the archives and ask for programming if your project needs. Similar to Landsat or SPOT data.

Spectral resolution: 4 bands multispectral (blue, green, red, NIR, MIR), 1 Pan band

4.3 Fine resolution (sub-meter to 20 m)

4.3.1 IKONOS from Space Imaging

4.3.2 Quickbird from Space Imaging

4.3.3 Orbview

5.0 Hyperspectral Data

5.1 Hyperion data

The first hyperspectral sensor on space. Can be exciting data for research. Cheap as well for a focused study area.

Spectral resolution: 220 bands multispectral (400 to 2500 nm, 10 nm wide continuous bands), 1 Pan

6.0 Radar data

6.1 JERS SAR data

6.2 Radarsat data

How to do a Catalogue of your Data Search

Ranjith, A.D., Biradar, C., Thenkabail, P.S.

An Example of a Data search for Nicolas Roost

1. Landsat -7 ETM+ Image ( Path: 124, row: 035; outer thick black lines) of the study area

2. ASTER Images of the study area

Data source: http://glovis.usgs.gov/ImgViewer/

Satellite	Sensor	Path/Row	Image Tile	Date of Acquisition	Quality Cloud
Quick bird	Quick-bird		1/6	2002-11-08	1%
			2/6	2002-06-22	14%
			3/6	2002-07-10	1%
			4/6	2004-02-14	0%
			5/6	2002-06-22	13%
			6/6	2002-07-10	0%

Satellite	Sensor	Path/Row	Image Tile	Date of Acquisition	Quality Cloud	Pixel size	Bands	Browse	Cost
Landsat-7	ETM+	124/39	1/6	1999-09-10	0%	30, 15	7+1	yes	600US$
				2000-10-30	0%	30, 15	7+1	yes	600US$
				2001-09-15	0%	30, 15	7+1	yes	600US$
				2002-01-05	0%	30, 15	7+1	yes	600US$
				2002-11-05	0%	30, 15	7+1	yes	600US$
				2003-03-29	0%	30, 15	7+1	yes	600US$

Satellite	Sensor	Path/Row	Image Tile	Date of Acquisition	Quality Cloud	Bands	Pixel size	Browse	Cost
EOS	Aster	124/039	1/4	2000-10-30	2%	4	MX15	yes	500US$
		124/039	2/4	2000-10-30	2%	4	MX15	yes	500US$
		124/039	3/4	2002-04-18	3%	4	MX15	yes	500US$
		124/039	4/4	2002-08-08	0%	4	MX15	yes	500US$

Satellite	Sensor	Path/Row	Image Tile	Date of Acquisition	Quality Cloud Cover1	Cost(Euro)
Spot-5	HRV	277/288	1/4	2004-05-17	BBBAAABA	? 8,100
		278/288	2/4	2004-06-22	AAAAAAAA	? 8,100
		278/289	3/4	2004-06-22	AAAAAAAA	? 8,100
		277/289	4/4	2004-05-17	BBBBBBAA	? 8,100

1.0 Overview

2.0 Where do you begin ?

3.0 How do you look for image/s you want ?

4.0 What types of satellite images suite your interest ?

4.0.1 Your first gateway for satellite images!

4.1 Coarse resolution (100 m and up)

4.1.1 AVHRR pathfinder

4.1.2 SPOT Vegetation

4.1.3 MODIS Terra and Aqua

4.2 Medium resolution (20-100 m)

4.2.1 Landsat data of 1970s, 1990s, and 2000s: Historical medium resolution data:

4.2.2 Landsat Geocover 1990s, and 2000s: Historical medium resolution data

4.2.3 Landsat data of recent times and historical: Historical medium resolution data

4.2.4 Landsat data of recent times and historical from USGS: Historical medium resolution data

4.2.5 Landsat (recent and historical) data from commercial sources of Spaceimaging: Medium resolution data

4.2.6 SPOT HRV data: Historical and recent

4.2.7 ASTER data

4.2.8 ALI data

4.2.9 IRS LISS data: Historical and recent

4.3 Fine resolution (sub-meter to 20 m)

4.3.1 IKONOS from Space Imaging

4.3.2 Quickbird from Space Imaging

4.3.3 Orbview

5.0 Hyperspectral Data

5.1 Hyperion data

6.0 Radar data

6.1 JERS SAR data

6.2 Radarsat data

How to do a Catalogue of your Data Search

Ranjith, A.D., Biradar, C., Thenkabail, P.S.

An Example of a Data search for Nicolas Roost

1. Landsat -7 ETM+ Image ( Path: 124, row: 035; outer thick black lines) of the study area

2. ASTER Images of the study area

Data source: http://glovis.usgs.gov/ImgViewer/

3. SPOT Images

Data Source: http://www.spotimage.com.cn/index.php

4. Quick-bird Images, for study area

Data Source: http://archive.digitalglobe.com

5. IKONOS Images of the study area