Dataset

Global PALSAR-2/PALSAR/JERS-1 Mosaics and Forest/Non-Forest Maps

1. Updates

Sep. 21, 2022
Replaced 33 tiles of the 2020 PALSAR-2 global mosaic datasets and updated to Ver.2.1.2.
The replaced tiles are as follows:

N70E165, N70E166, N70E167, N70E168, N70E169, N70E170, N70E171, N70E172, N70E173, N70E174, N70E175, N70E176, N70E177, N70E178, N70E179, N71E165, N71E166, N71E167, N71E168, N71E169, N71E170, N71E171, N71E172, N71E173, N71E174, N71E175, N71E176, N71E177, N71E178, N71E179, N72E177, N72E178, N72E179

Jul. 22, 2022
PALSAR-2 global mosaic datasets from 2015 to 2021 were updated to Ver.2.1.1.

Global 25m Resolution PALSAR-2 Mosaic Dataset Description has been updated (Ver. 2.1.1).

Jun. 21, 2022

PALSAR-2 global mosaic datasets from 2015 to 2021 were updated to Ver.2.1.0

Global 25m Resolution PALSAR-2 Mosaic Dataset Description has been updated (Ver. 2.1.0).

Global 25m Resolution PALSAR-2 Global Forest/Non-forest Map Dataset Description has been updated (Ver. 2.0.0a).

Global 25m Resolution PALSAR Mosaic and Forest/Non-Forest Map (FNF) Dataset Description has been updated (Ver. O).

Apr. 12, 2022

PALSAR-2 global mosaic datasets in 2015 and 2016 were updated, and 2021 was released.

Forest/non-forest maps in 2017, 2018, 2019 and 2020 were released.

The dataset description document was updated and divided into the global mosaic dataset and the forest/non-forest map dataset.

Mar. 22, 2022

PALSAR-2 global mosaic datasets in 2017 and 2018 were updated.
The dataset description document has been updated (Ver. M).
Forest/non-forest maps in the Year 2017 and 2018 is currently in preparation for release.

Feb. 28, 2022

PALSAR-2 global mosaic map in 2019 and 2020 were updated.

The dataset description document has been updated (Ver. L).

Forest/non-forest maps are not included in 2019 and 2020 dataset, yet(Coming soon).
Apr. 08, 2021

PALSAR-2 global mosaic map in 2020 was released.

The dataset description document has been updated (Ver. K).

Forest/non-forest map is not included in 2020 dataset, yet.

Mar. 31, 2021

PALSAR-2 global mosaic map in 2019 was released.

The dataset description document has been updated (Ver. J).

Forest/non-forest map is not included in 2019 dataset, yet.

Dec. 27, 2019

PALSAR-2 global mosaic map in 2018 is released.

The dataset description document has been updated (Ver. I).

Forest/non-forest map is not included in 2018 datasets, yet.

May. 07, 2018 PALSAR-2 global mosaic and forest/non-forest map (2017 global dataset, and low resolution dataset of 2015-2017) are released.
The dataset description document has been updated (Ver. H).
Apr. 27, 2018 The dataset description document has been updated (Ver. G).
Apr. 19, 2018 PALSAR-2 global mosaic and forest/non-forest maps (2015 and 2016 datasets around Japan) are released.
The dataset description document has been updated (Ver. F).
Oct. 2, 2017 The dataset description document has been updated (Ver. E).
Jun. 20, 2017 Replaced 9 tiles of the 2015 dataset. The replaced tiles are as follows:
N16E145,N16E146,N17E145,N17E146,N18E145,N18E146,N19E145,N19E146,N20E145
May. 24, 2017 About data utilization has been added.
Apr. 27, 2017: PALSAR-2 global mosaic and forest/non-forest map in 2016 are released. Reprocessed version of PALSAR-2 global mosaic and forest/non-forest map in 2015 are released. Dataset description document has been updated (Ver.D).
* The data around Japan will be available in the near future.
Jan. 11, 2017 JERS-1 SAR yearly mosaic for tropical regions from 1993 to 1998 are released. The dataset description document has been updated (Ver. C).
Oct. 31, 2016 JERS-1 SAR global mosaic in 1996 are released. Dataset description document has been updated (Ver. B).
May. 10, 2016 Dataset description document has been updated (Ver. A).
Apr. 6, 2016 (Correction) of figure 1 used in [Global 25m Resolution PALSAR-2/PALSAR Mosaic and Forest/Non-Forest Map]
Mar. 28, 2016 PALSAR-2 global mosaic and forest/non-forest map for Japan in 2015 are released.
Mar. 28, 2016 Low resolution forest/non-forest maps (0.25 degrees and 1km) are released.
Jan. 25, 2016 PALSAR-2 global mosaic and forest/non-forest map in 2015.
* Data for Japan (using polarimetry mode) will be available soon.
Jan. 16, 2014 PALSAR global mosaic and forest/non-forest maps in 2007, 2008, 2009 and 2010.

2. Overview

To improve understanding of and responses to the impacts of global environmental changes, timely and consistent monitoring of the global land cover is essential. Global monitoring with satellite remote sensing is one of the most effective approaches to detect land surface changes, such as deforestation and forest degradation, as satellites can provide wall-to-wall observations covering wide areas periodically. The long wavelength (L-band) Synthetic Aperture Radars (SAR) on ALOS and ALOS-2 are particularly sensitive to vegetation structure and can observe the land surface even under clouds, and as such, are particularly useful to provide information about forest changes in the tropical regions.

Global L-band SAR mosaics

The global 25m resolution SAR mosaics and forest/non-forest maps are free and open annual datasets generated by JAXA using the L-band Synthetic Aperture Radar sensors on the Advanced Land Observing Satellite-2 (ALOS-2 PALSAR-2), the Advanced Land Observing Satellite (ALOS PALSAR) and the Japanese Earth Resources Satellite-1 (JERS-1 SAR).

The mosaics have been created by assembling long paths of SAR backscatter images observed through JAXA’s global Basic Observation Scenario for PALSAR and PALSAR-2. Correction of geometric distortions specific to SAR (ortho-rectification) and topographic effects on image intensity (radiometric slope correction) are applied to accommodate image analysis without the need for further processing. The mosaics are provided as Gamma-0 backscatter, in geographical (lat/long) coordinates with a pixel spacing of 0.8 arc seconds (approximately 25 meters at the Equator). The temporal interval of the mosaics is annual with PALSAR-2 mosaics available for the 2015-2021 time period and PALSAR mosaics for 2007-2010. The JERS-1 SAR mosaic dataset is also available for 1993-1998.

In 2022, a set of re-processed global 25m resolution PALSAR-2 mosaics for the period 2015 to 2020 (Ver. 2) were released together with a newly created global PALSAR-2 mosaic for the year 2021. The new PALSAR-2 mosaics have improved geometric performance and radiometric balancing between adjacent paths to make the products as seamless as possible. The differences in processing methods and output formats compared to previous versions are as follows;

  • GDAL based mosaicking process to improve absolute geometric accuracy and processing speed.
  • Products developed to comply with CEOS Analysis Ready Data for Land, Normalised Radar Backscatter (CARD4L NRB) specifications.
  • Only data from the one year in question are used for each annual mosaic, and hence no gap-filling using data from previous years in case of gaps in the annual global coverage.
  • Radiometric balancing between adjacent paths is applied to land data only.
Due to the above changes in the mosaicking approach, please be aware of the following possible phenomena when comparing the new PALSAR-2 datasets with PALSAR and JERS-1 SAR mosaic datasets;
  • The backscatter may differ from pass to pass in the ocean due to the lack of radiometric balancing, and in some land areas due to changes in backscattering intensity caused by changes in land cover, soil moisture, or seasonal variations.
  • Due to the geometric accuracy improvement, slight misalignment may be observed when the new global 25m resolutions PALSAR-2 mosaic dataset (Ver.2) is overlaid on the past PALSAR-2/PALSAR/JERS-1 mosaic datasets.

Re-processing of the annual (2007-2010) PALSAR mosaic datasets in accordance with the new approach and product format is planned to follow.

For details on the re-processed PALSAR-2 datasets, please see "Global 25m Resolution PALSAR-2 Mosaic Dataset Description". For the global 25m resolution PALSAR datasets, please see the "Global 25m Resolution PALSAR Mosaic and Forest/Non-Forest Map (FNF) Dataset Description".

Global Forest/Non-Forest maps

The global forest/non-forest maps (FNF) were generated by a Random Forest machine learning-based classification method, with the re-processed global 25m resolution PALSAR-2 mosaic dataset (Ver. 2.0.0) as input. Here, the "forest" is defined as the tree covered land with an area larger than 0.5 ha and a canopy cover of over 10 %, in accordance with the FAO definition of forest. The classification results are presented in four categories, with two categories of forest areas: forests with a canopy cover of 90 % or more and forests with a canopy cover of 10 % to 90 %, depending on the density of the forest area.

The previous version of the global forest/non-forest maps - currently the case for the PALSAR FNF maps - were published in three categories: forest, non-forest, and water. A threshold method was used to determine forest/non-forest for each year, which caused discontinuities when examining the same area over time, and changes could not be accurately captured.

Therefore, in the newly released (Ver 2.0.0) version, data for the past year is used in addition to the data for the target year to be created. For example, when creating the 2020 FNF, the global 25m resolution mosaics of 2019 and 2020 are stacked and used as input data for classification. When stacked, the target year mosaic is weighted 3 times more than the past year's mosaic. In this way, the stacked data can represent the forest change by retaining the effect of the target year on the areas where the forest has changed.

Accuracy validation of the classification results was conducted for the four-category case and for the three-category case where the two forest areas were integrated. The results indicate that the overall accuracy was at least 86 % for 3 categories and 76 % for 4 categories for four years from 2017 to 2020. Accuracy thus improved by 5-10 % or more compared to the previous version. Figure 1 shows global mosaic images and forest/non-forest maps, and Figure 2 shows an example of Borneo island. Please refer to the "Global 25m Resolution PALSAR-2 Forest/Non-forest Map Dataset Description" document for details on the classification method.
JAXA plans to update the FNF classification algorithm to improve its accuracy. Any comments or requests for the datasets are welcome.

Figure 1: PALSAR-2 global mosaics and forest/non-forest maps.

Figure 2: Loss of forest cover (red areas in the map) in Borneo island during 2017 to 2019.

Figure 2: Loss of forest cover (red areas in the map) in Borneo Island from 2017 to 2019.

3. Dataset description

SAR processing algorithm:
  • ALOS-2 PALSAR-2: Sigma-SAR IMAGE (for strip processing) and GDAL (for mosaicking)
  • ALOS PALSAR and JERS-1 SAR: Sigma-SAR IMAGE (for strip processing) and Sigma-SAR MOSAIC (for mosaicking)
Data format:
  • Datasets for 2015 and beyond: Cloud Optimized GeoTIFF + XML metadata (in compliance with CARD4L NRB (CEOS Analysis Ready Data for Land, Normalised Radar Backscatter Ver.5.5) specifications)
  • Datasets up to 2010: binary image file + header file
Pixel spacing and data size:
  • 0.8 arc seconds (approx. 25 m at the Equator)
  • 1 x 1 degree tiles (4500 x 4500 pixels)
Dataset time periods:
  • ALOS-2 PALSAR-2 mosaics and FNF maps: 2015, 2016, 2017, 2018, 2019, 2020, 2021
  • ALOS PALSAR mosaics and FNF maps : 2007, 2008, 2009, 2010
  • JERS-1 SAR mosaics: 1993, 1994, 1995, 1996, 1997, 1998
Contents:
  • Ortho and slope corrected Gamma-0 backscattering coefficient (ALOS-2 PALSAR-2 and ALOS PALSAR: HH and HV polarization, JERS-1 SAR: HH polarization)
  • Forest/non-forest map
  • Observation date image
  • Local incidence angle image
  • Processing mask information image

For details, please see the following dataset descriptions;

25m global mosaics and forest/non-forest maps up to 2010 observed by ALOS PALSAR sensor:
25m global mosaics from 2015 to 2021 observed by ALOS-2 PALSAR-2 sensor:
Forest/non-forest maps using ALOS-2 PALSAR-2 global mosaic from 2017 to 2020

4. Download

Please register your information from the following URL to download the dataset. Your email address is required.

This is a temporary registration (please confirm your email address). A confirmation email will be sent to you to accept your registration request. After the confirmation of your request, the download information will send by a separate email.
The dataset download site is
If you forget your password, please register again.
Microsoft Edge, Google Chrome, Mozilla Firefox, or Apple Safari are recommended.
Depending on your network environment, the downloaded compressed file may be automatically uncompressed.
Depending on the time of the day, it may take time to display the page or download the datasets.

5. About data utilization

About decompression of data

After login to the download page, use the global map to zoom into your geographical area of interest. Clicking on the location of a latitude / longitude grid cell on the map on the screen display makes it possible to download the datasets in 1 × 1 degree tiles or 5 × 5 degree units.
The downloaded file is compressed into one file by zip or tar.gz and software may be necessary for this decompression. For example, 7-Zip is available as free software.

Displaying and using data

To display and use the "Global PALSAR-2/PALSAR/JERS-1 Mosaic and Forest/Non-Forest map", image processing, remote sensing or GIS software are required.

Public open GIS and image processing software that can be used include, for example,

Please note that the software above are mentioned as examples only, and their utility cannot be guaranteed by JAXA. Please understand that JAXA can not answer any questions about the software.

Please refer to these as "JAXA Global PALSAR-2/PALSAR/JERS-1 Mosaic and Forest/Non-Forest maps".

6. Note for data use

This dataset is available to use with no charge under the "Terms of Use of Research Data".
To publish images using the dataset as it is, please display "©JAXA" in the image.

7. Contact

ALOS-2/ALOS Science Project

Earth Observation Research Center (EORC)

Japan Aerospace Exploration Agency (JAXA)

E-mail: aprojectjaxa.jp

Please refer to FAQ before sending an email.
Please contact us in English or Japanese.

8. References

9. Old Version

PALSAR 10m Mosaic Data
PALSAR mosaic dataset with 10m spatial resolution.
  • Porto Velho, Rondonia, Brazil
  • Riau, Indonesia
  • Congo River
Global Forest / Non-forest map
The global forest and non-forest area distribution in 2007 and 2009 by PALSAR.