High-Resolution Land-Use and Land-Cover Map of Vietnam for 2020
(Released in September 2023 / Version 23.09) 

 You can switch layers with check boxes.

Download product

User registration is required to download the data. Please click here about download and user registration details.

1. Summary

Japan Aerospace Exploration Agency (JAXA) Earth Observation Research Center (EORC) has released several High-Resolution LULC products for Mainland Vietnam (HRLULC-Vietnam) to provide basic information for a variety of applications for the country such as biodiversity evaluation, natural disaster countermeasures, urban planning, etc. This time, JAXA has released HRLULC-Vietnam "version 23.09", which reflects the LULC status over Mainland Vietnam (excluding some remote islands in the east) as of the year 2020 with a spatial resolution of 10 m. We aimed to increase classification accuracy by using time series satellite data in combination with the CNN approach developed in the High-Resolution LULC mapping of Japan (HRLULC-Japan v21.11). We succeeded in creating a 12-category LULC for Vietnam with very high accuracy (see the accuracy verification section), which is much higher than the previous 10-m LULC maps for Vietnam. The LULC classification system was inherited from the previous studies. The main differences and improvements compared with the previous HRLULC-Vietnam with a 10-m spatial resolution (v20.06 and v21.04) are as follows:
  • Higher overall accuracy
  • Used time series of satellite data (Sentinel-2, Sentinel-1, PALSAR-2/ScanSAR with 6 seasons)
  • Applied CNN approach which spans over time and features dimensions
  • A better spatial detail compared to previous versions.
  • Added Aquaculture category and remove Evergreen Needleleaf Forest category.

2. Data used for classification

  • Data 1: Sentinel-2 Level-2A (*1)
  • Data 2: Sentinel-1 Level 1 (*1)
  • Data 3: ALOS-2/PALSAR-2/ScanSAR Level 2.2
  • Data 4: ALOS/PRISM/DSM
  • Data 5: OpenStreetMap (road networks) (*2)

3. Classification algorithm

3.1 Preprocessing

  • Sentinel-2: masking cloud, creating median composite using multiple years data, creating vegetation indices (NDVI, EVI, GNDVI, GRVI, NDWI, GSI)
  • Sentinel-1: angular-based radiometric correction, speckle noise filtering, creating indices (VH-VV, VH/VV)
  • PALSAR-2/ScanSAR: speckle noise filtering, creating indices (HH-HV, HH/HV, NDI, NLI)
    • *NDVI: Normalized Difference Vegetation Index, EVI: Enhanced Vegetation Index, GNDVI: green normalized difference vegetation index, GRVI: Green and Red ratio Vegetation Index, NDWI: Normalized Difference Water Index, GSI: Grain Size Index, NDI: Normalized Difference Index.

3.2 Algorithm

Convolutional Neural Network classification with convolutional process was conducted over time and feature dimension. The source code of the algorithm was inherited from the SACLASS-2 with some modifications to compatibility with 6-season data.

4. Data format

Coordinate system Latitude and longitude coordinate system with WGS84
Tile unit 1 degree x 1 degree, (11,133 pixels x 11,133 lines)
Mesh size (1 / 11,133) degree × (1 / 11,133) degree (corresponding to approx. 10 m × 10 m)
File naming convention For example, N08E104_2020_v23.09_10m.tif:
- N08E104: indicates 8 to 9 degrees north latitude and 104 to 105 degrees east longitude
- 2020: indicates the year of the LULC map
- v23.09: indicates the version of the product, including the year (2023) and the month (09-September), in which the map was published
- 10m: indicates the spatial resolution of the map is 10 meters
Format GeoTIFF format
Period of coverage Year 2020 (not a specific point of time but the average situation; some satellite data in 2019 was used for missing data in 2020)

The value of each pixel is the ID number of the category for classification as follows:

  • #1: Water
  • #2: Urban/Built-up
  • #3: Rice
  • #4: Other Crops
  • #5: Grass/Shrub
  • #6: Woody Crops/Orchards
  • #7: Barren
  • #8: Evergreen Forest
  • #9: Deciduous Forest
  • #10: Plantation Forest
  • #11: Mangrove Forest
  • #12: Aquaculture

5. Accuracy verification

Validation data were collected using the stratified random sampling method, following the good practice of accuracy assessment for LULC change proposed by Olofsson et al. (2014). 600 points were distributed randomly on 12 strata corresponding to 12 LULC categories. An error matrix of sample count (Table 1) was created based on the validation data. After that, the error matrix of area proportion was calculated by multiplying each row of Table 1 with the map’s area proportion of the category in that row. The overall accuracy, user’s accuracy, producer’s accuracy, and their standard deviation were calculated directly from Table 2.
Table 1: Error matrix of sample count
Table 1: Error matrix of sample count
Table 2: Error matrix of area proportion
Table 2: Error matrix of area proportion
Figure 1: Resultant LULC map for mainland Vietnam in 2020.
Figure 1: Resultant LULC map for mainland Vietnam in 2020. (a) LULC map for mainland Vietnam with 12 categories, (b) a zoom-in area of the LULC map in (a) for Hanoi City and the surrounding area, (c) a zoom-in area of the LULC map in (a) for Ho Chi Minh City and the surrounding area.
Figure 2: The urban expansion
Figure 2: The urban expansion in Ho Chi Minh City during 2016 and 2020 by comparing between version 20.06 (a) and version 23.09 (b)
Figure 3: The improvement
Figure 3: The improvement of version 23.09 (b) compared to version 20.06 (a) for the classification of LULC in Tan Son Nhat airport in Ho Chi Minh City.
Figure 4: Land cover change from grassland to solar panel (Nihonmatsu City, Fukushima prefecture)
Figure 4: The aquaculture category was added in version 23.09 (b). Version 20.06 (a) classified most of aquaculture areas into rice or water

6. References

  • Sota HIRAYAMA, Takeo TADONO, Masato OHKI, Yousei MIZUKAMI, Kenlo NISHIDA NASAHARA, Koichi IMAMURA, Naoyoshi HIRADE, Fumi OHGUSHI, Masanori DOTSU and Tsutomu YAMANOKUCHI (2022): Generation of High-Resolution Land Use and Land Cover Maps in JAPAN Version 21.11. Journal of The Remote Sensing Society of Japan Vol. 42 No. 3 pp. 199-216
  • Olofsson, P. et al. (2014): Good practices for estimating area and assessing accuracy of land change. Remote. sensing Environ. 148, 42-57.