What is new?

This system can provide the most recent picture of forest status over large regions, free of clouds and consistently updated on a regular basis (monthly to yearly)

Example deforestation animation

NB: to view the deforestation in Google^TM Earth, scroll down a bit.

Example results of the regional resolution monitoring system. This particular version shows forest (green), non forest (gray), deforestation since 1998 (red), and recent deforestation during the preceding 3-months (yellow). Time series like these clearly depict conversion trends. (Note: cumulative deforestation since 1998 is depicted, regrowth is not considered in this example).

View deforestation maps in Google^TM Earth

Interested in viewing example maps in Google^TM Earth yourself?
Clicking on the links below opens maps automatically in Google^TM Earth for easy visualisation (note it may take a while loading these high resolution images!!).

1. Deforestation time-series "movie" Sumatra 1998 - 2007 (July)

This map series shows cumulative deforestation since 1998 on a yearly basis. As an example, only Sumatra has been selected. Note this is a first prototype of the system developed in collaboration with the Indonesian Ministry of Forestry. Contact us for a higher resolution improved version.

Instructions to play the animation in Google^TM Earth:
After clicking the link below, it will load automatically in Google^TM Earth (you'll need to have Google^TM Earth installed on your pc, see below where to get it). After loading the data, a timeslider will appear in the upper right corner of your Google^TM Earth window, next to the compass. (You'll need a pretty recent version of Google^TM Earth for this, so if the time slider doesn't appear, your version of Google^TM Earth might be outdated; in the 'help' menu sits a function to check for updates).
To start the animation, click on the ‘play' button (the little triangle) right of the slider. Loading the first time might take some time, a second run of the animation will run smoother. Left of the time slider you'll find a little clock icon, click on that to adjust animation settings like speed and wrapping around or not.

Of course you can zoom in to show the animation for a smaller area only, with more detail. Because higher detail data needs to be loaded, again the first run might not be as smooth as possible. Just run it a second time, the data will be in cache now and the animation will run smoothly.

Click here to open the deforestation map series in Google^TM Earth

2. Forest status 2000 - 2007 (July) of insular South East Asia (static image, no animation)

This map shows a recent update of forest status over entire insular South East Asia, including the location of peat swamp forests critical to global climate and deforestation since the 2000. Note this is a first prototype of the system, contact us for a higher resolution improved version.

Click here to open the forest status map in Google^TM Earth

If you do not have the latest (version 4.2) Google^TM Earth version installed, download or upgrade it here for free: http://earth.google.com/download-earth.html

Field photos used to assess deforestation in locations where changes have been detected by satellite. Riau province, Sumatra. Photos courtesy WWF/KWS

Background

The regional resolution monitoring system, the first of its kind in operation, allows for the development of consistent time series of accurate land use / cover and land use / cover change data. The system is optimised for large scale cloud cover independent monitoring, providing:

• Economically feasible information by using freely available or cheap satellite data;

• Permanent public service to provide up-to-date information continuously on forest status and change, through the Internet (e.g. Google^TM Earth);
• Full coverage at the regional scale (e.g. S.E Asia, Guiana Shield and Congo basin);

• Timeliness; new status and change map updates should be ready for use fast: i.e. every 3 months instead of two to five years;

• Easy integration with GIS data sources (e.g. rivers, concession borders, socio-economic data) in order to do statistical calculations and further analyses;

• An alarm system to find change hotspot areas to inform (international) policy and decision makers and assess in more detail using higher resolution systems;

The system uses hundreds of freely available satellite images from the MODIS, SPOT Vegetation and ALOS PALSAR radar satellites, complemented with large Envisat Wide Swath radar images. It provides area-wide information at a resolution of 50 - 1000 m. The main advantage of MODIS and SPOT sensors is that they observe the globe several times a day. This allows us to select and put all the cloud free parts of daily images together to derive clear pictures. The PALSAR and Envisat ASAR radar sensors see through clouds, haze and smoke to make clear pictures.

Processing

An automated image processing method has been developed to adjust for atmospheric disturbance, effectively eliminate clouds, smoke, and haze by combining data from many raw (radar) input images and enhance contrast between vegetation types. Using this approach, a sufficiently cloud free image suitable for monitoring, classification and change detection is available every 3 months.

Example of a raw cloud free image aggregate of SPOT VEGETATION satellite images, in this case for the period April 1998-March 1999. Such data is used as input for both the land cover/use maps and the land cover/use change map products.

By analyzing the complete sequence of 3-month images since 1998 (or any other year), the probability of a pixel belonging to a particular land use / cover class is calculated, also taking into account surrounding pixels.

Land cover / use map of Indonesia for March 1999 developed from the raw satellite imagery (note: this information is updated and made available every 3 months).

Detail of land cover / use map of Indonesia for March 1999 (note: this information is updated and made available every 3 months).

Land use / cover change maps are derived from automatic analysis of changes between different dates.

Detail of land cover / use change map of Indonesia for 1998-1999 (note: this information is updated and made available every 3 months).

Application fields

For informed decision making it is important to have a fast, timely overview of the status of forest due to recurrent fire impacts, rapidly expanding oil palm plantations, illegal logging etc. Employing systematic, permanent monitoring using time-series can increase insight in the magnitude of recent or current environmental problems and promote awareness about these problems. Besides, it can provide early warning signs of potential future environmental problems by closely tracking change dynamics. The main areas of recent change can easily be isolated and further investigated by higher resolution sensors, e.g. spaceborne radar (ASAR, PALSAR) and traditional sensors (Landsat, ASTER, SPOT HR) or even airborne radar systems.

This information can be linked to GIS data sources in order to provide maps and accompanying statistical data for regular policy briefings.

Examples

• A prototype of the regional system using MODIS/SPOT Vegetation satellite data is used by the Indonesian Ministry of Forestry to assess deforestation trends. See Press Release Jakarta Post.

• Work is ongoing with several governments in the Guiana Shield area in South America (i.e. Colombia) and West and Central Africa.

 

SarVision is not an official source of information from Google that produces and owns Google^TM Earth. Google and Google Earth are trademarks of Google Inc. Original image screen shots from Google^TM Earth are Copyright Google.

---