10

I have just discovered amazing Hansen´s classification about forest loss and re-growth, available at http://www.globalforestwatch.org/, published in Science, 2013 as: Hansen, M. C., Potapov, P. V, Moore, R., Hancher, M., Turubanova, S. A., & Tyukavina, A. (2013). High-Resolution Global Maps of 21st-Century Forest Cover Change. Science, 342(6160)(15 November), 850–854. doi:DOI:10.1126/science.1244693. enter image description here

However, I can´t find in this article/on the website the exact methodology how to reproduce such a map so which classification has Hansen used?

The only thing I can find is that the supervised learning algorith was used to identify tree cover, but it is quite a broad term.

If it is possible, I would like to use the same methodology (but apply it on 90th years), so before Hansen´s classification in my selected area.

  • I think that his code (at leat part of it) is available on Google earth engine earthengine.google.org/#intro – radouxju Jan 18 '15 at 19:52
  • 1
    try to contact the authors: mhansen@umd.edu – julien Jan 20 '15 at 15:31
  • I was looking into this as well to try and downscale and replicate to a small region and compare accuracy vs. official government data. Like @Mikkel, the best I could find was the supplementary paper. It seems the exact methodology was never publicly published (?). It is worth contacting Dr. Hansen. However, given that the methodology has not been published, it seems unlikely that he would specify this now. The Forest Watch team and Hansen are actively working on improving the analysis and correcting errors, so they may not want to release the algorithm in its current state. – Dan Jan 21 '15 at 8:52
6

Matt Hansen's team has a paper published on forest cover change in Eastern Europe that goes back to 1985 - see Eastern Europe's forest cover dynamics from 1985 to 2012 quantified from the full Landsat archive http://www.sciencedirect.com/science/article/pii/S0034425714004817

I'm also checking with colleagues on whether Matt Hansen's algorithm is available for use within Google Earth Engine.

In the meantime, we'll be updating the Hansen dataset on Global Forest Watch in February, to include data through 2013.

8

The Supplementary Materials (SM) for the Science article provides references to a number of different journal-articles that outline various parts of the methodology.

The SM can be found here

Extending the time-series to include Landsat-5 (and potentially Landsat-8 to make the methodology something that can be rerun "easily") data will be a challenging task, and will require extensive testing of the image normalization. The image normalization may be made even more difficult, due to the lack of coinciding MODIS coverages. Instead, you may have to apply a different normalization approach, such as a combination of COST - article and MAD - article as outlined in this article.

All-in-all, it is not a simple task, and I wish you good luck with it.

The most relevant articles from the SM are:

P. Potapov, S. A. Turubanova, M. C. Hansen, B. Adusei, M. Broich, A. Altstatt, L. Mane, C. O. Justice, Quantifying forest cover loss in Democratic Republic of the Congo, 2000- 2010. Remote Sens. Environ. 122, 106–116 (2012). Article

M. Broich, M. C. Hansen, P. Potapov, B. Adusei, E. Lindquist, S. V. Stehman, Time-series analysis of multi-resolution optical imagery for quantifying forest cover loss in Sumatra and Kalimantan, Indonesia. Int. J. Appl. Earth Obs 13, 277–291 (2011). Article

M. Hansen, A. Egorov, D. P. Roy, P. Potapov, J. Ju, S. Turubanova, I. Kommareddy, T. R. Loveland, Continuous fields of land cover for the conterminous United States using Landsat data: First results from the Web-Enabled Landsat Data (WELD) project. Remote Sens. Letters 2, 279–288 (2011). Article

M. Hansen, R. S. DeFries, J. R. G. Townshend, M. Carroll, C. Dimiceli, R. A. Sohlberg, Global percent tree cover at a spatial resolution of 500 meters: First results of the MODIS vegetation continuous fields algorithm. Earth Interact. 7, 1–15 (2003). Article

L. Breiman, J. Friedman, R. Olsen, C. Stone, Classification and Regression Trees Wadsworth and Brooks/Cole, Monterey, CA, (1984).

  • Here's the thing. The Landsat-missions are supposed to be used for change monitoring, but the sensor characteristics are so different that comparing different Landsat data is very difficult. For me this decision to build the system this way is really hard to understand. We should have something that is constistent. If you want to add something new, do it, but keep the continuum with previous instruments. I really hope that Landsat 9 will have the exact same characteristics as Landsat 8 OLI (+ maybe some extra bands :) ) – reima Oct 16 '18 at 3:30

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.