1 / 47

MCD45 Burned Area (500m approximate day of burning)

MCD45 Burned Area (500m approximate day of burning). David Roy*, Luigi Boschetti # , Chris Justice # *South Dakota State University, GIS Center of Excellence # University of Maryland, Department of Geography. Overview. Algorithm Background MCD45 Product Suite Product Examples

kim
Télécharger la présentation

MCD45 Burned Area (500m approximate day of burning)

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. MCD45 Burned Area(500m approximate day of burning) David Roy*, Luigi Boschetti#, Chris Justice# *South Dakota State University, GIS Center of Excellence # University of Maryland, Department of Geography MCD45 Global Burned Area Product

  2. Overview • Algorithm Background • MCD45 Product Suite • Product Examples • Product Data Structure • Validation • Collection 5 Recommendations MCD45 Global Burned Area Product

  3. Algorithm • New approach - Rolling BRDF based expectation change detection • Semi-Physically based; less dependent upon imprecise but noise tolerant classification techniques • Automated, without training data or human intervention • Applied independently per pixel to daily gridded MODIS 500m land surface reflectance time series => map 500m location and approximate day of burning Algorithm Background MCD45 Global Burned Area Product

  4. The challenge: change detection of Burned Areas BRDF Effects gaps Algorithm Background MCD45 Global Burned Area Product

  5. The challenge: change detection of Burned Areas Persistence of the signal BRDF Effects gaps Day of burning Algorithm Background MCD45 Global Burned Area Product

  6. Conceptual Scheme observed r time Algorithm Background MCD45 Global Burned Area Product

  7. Conceptual Scheme observed r t-1 time Algorithm Background MCD45 Global Burned Area Product

  8. Conceptual Scheme observed BRDF Inversion window r t-1 time Algorithm Background MCD45 Global Burned Area Product

  9. Conceptual Scheme observed BRDF Inversion window predicted r > r (t|t-1) t-1 time Algorithm Background MCD45 Global Burned Area Product

  10. Conceptual Scheme observed BRDF Inversion window predicted r > r (t|t-1) r (t|t-1) t-1 time Algorithm Background MCD45 Global Burned Area Product

  11. Conceptual Scheme observed BRDF Inversion window predicted r > r (t+1|t) r (t+1|t) t time Algorithm Background MCD45 Global Burned Area Product

  12. …looking at one algorithm step: Predicted Reflectance Algorithm Background MCD45 Global Burned Area Product

  13. …looking at one algorithm step: Predicted Reflectance Observed Reflectance Algorithm Background MCD45 Global Burned Area Product

  14. …looking at one algorithm step: Predicted Reflectance Observed Reflectance Algorithm Background MCD45 Global Burned Area Product

  15. …looking at one algorithm step: Predicted Reflectance probability of change: Z-score = (predicted-observed)/error Observed Reflectance Algorithm Background MCD45 Global Burned Area Product

  16. Conceptual Scheme observed BRDF Inversion window predicted Consistency window r > r (t+1|t) r (t+1|t) t time first detection Algorithm Background MCD45 Global Burned Area Product

  17. Conceptual Scheme observed BRDF Inversion window predicted Consistency window r > r (t+2|t) r (t+2|t) t time first detection Algorithm Background MCD45 Global Burned Area Product

  18. Conceptual Scheme observed BRDF Inversion window predicted Consistency window r > r (t+3|t) r (t+3|t) t time first detection Algorithm Background MCD45 Global Burned Area Product

  19. Conceptual Scheme observed BRDF Inversion window predicted Consistency window r > r (t+4|t) r (t+4|t) t time first detection Algorithm Background MCD45 Global Burned Area Product

  20. The MCD45 product suite • Monthly Product (MCD45A1) • 500m approximate day of burning over month +/- 8 days • MODLAND 10x10 degree tiles • extensive QA & metadata • EOS-HDF • Annual Synthesis Product (forthcoming) • as monthly product but for calendar year • CMG 0.25 deg product (forthcoming) • aggregated information for climate modeling community (continental scale, easily accessible format) MCD45 Product Suite MCD45 Global Burned Area Product

  21. Collection 4 Protoyping Central Southern Africa 500m burned areas Product Example MCD45 Global Burned Area Product

  22. 500m burned areas 5 months 2002 Zambia/Zimbabwe 650*500km Product Example MCD45 Global Burned Area Product

  23. 1km active fires 5 months 2002 Zambia/Zimbabwe 650*500km Product Example MCD45 Global Burned Area Product

  24. Australia 500m burned areas 1 month 2002 Product Example MCD45 Global Burned Area Product

  25. Australia 1km active fires 1 month 2002 Product Example MCD45 Global Burned Area Product

  26. Russian Federation 500m burned areas 1 month 2002 Product Example MCD45 Global Burned Area Product

  27. Russian Federation 1km active fires 1 month 2002 Product Example MCD45 Global Burned Area Product

  28. Brazil, Southern Para, 500m burned areas 1 month 2002 Product Example MCD45 Global Burned Area Product

  29. Brazil, Southern Para, 1km active fires 1 month 2002 Product Example MCD45 Global Burned Area Product

  30. First Global Collection 5 ProductNovember 2000(shown in 40 x 40km grid cells) Product Example MCD45 Global Burned Area Product

  31. First Global Collection 5 ProductNovember 2000(shown in 10 x 10km grid cells) Product Example MCD45 Global Burned Area Product

  32. Product Data Structure For each 500m pixel: • “Burndate” the approximate day of burning • “BA pixel QA” confidence of the detection (high, moderate, low). • “Number Passed” number of observations in the time series after/before “Burndate” that were labeled as burned. • “Number Used” number of observations in the time series after/before “Burndate” that were available for consideration (“Number Used” ≥ “Number Passed”). • “Direction” direction in which burning was detected (forward in time, backward, or both). • “Surface Type” information describing the static land cover type, worst sensing acquisition and aerosol state conditions over the input time series. • “Gap Range” information describing the largest and the second largest number of missing/cloudy days (if any) in the time series + Tile level ECS, MODLAND and product specific metadata MCD45A1 Data Structure MCD45 Global Burned Area Product

  33. A closer look at the monthly product… Tile h17 v06 2000306 (November 2000) 205 251 SDS “Burndate” MCD45A1 Data Structure MCD45 Global Burned Area Product

  34. A closer look at the monthly product… 500m Burnt Areas 1km Active Fires Tile h17 v06 2000306 (November 2000) 205 251 SDS “Burndate” MCD45A1 Data Structure MCD45 Global Burned Area Product

  35. A closer look at the monthly product… Tile h17 v06 2000306 (November 2000) 205 251 SDS “Burndate” MCD45A1 Data Structure MCD45 Global Burned Area Product

  36. A closer look at the monthly product… 205 251 SDS “Burndate” MCD45A1 Data Structure MCD45 Global Burned Area Product

  37. A closer look at the monthly product… 1 2 3 4 SDS “BA Pixel QA” MCD45A1 Data Structure MCD45 Global Burned Area Product

  38. A closer look at the monthly product… 1 1(forward) 2 2(backward) 3 3(both) SDS “BA Pixel QA” SDS “Direction” MCD45A1 Data Structure MCD45 Global Burned Area Product

  39. A closer look at the monthly product… 277 365 SDS “gap_range1”: gap start MCD45A1 Data Structure MCD45 Global Burned Area Product

  40. A closer look at the monthly product… 1 16 SDS “gap_range1”: gap duration MCD45A1 Data Structure MCD45 Global Burned Area Product

  41. A closer look at the monthly product… 277 365 SDS “gap_range2”: gap start MCD45A1 Data Structure MCD45 Global Burned Area Product

  42. A closer look at the monthly product… 1 16 SDS “gap_range2”: gap duration MCD45A1 Data Structure MCD45 Global Burned Area Product

  43. Validation Plan • Currently CEOS stage 1 validation • “the product accuracy has been estimated using a small number of independent measurements obtained from selected locations and time periods and ground-truth/field program efforts” • Stage 2 by the end of Collection 5 • “the product accuracy has been assessed over a widely distributed set of locations and time periods via several groundtruth and validation efforts” • Stage 3 in Collection 6 • “product accuracy has been assessed and the uncertainties in the product well established via independent measurements in a systematic and statistically robust way” Validation MCD45 Global Burned Area Product

  44. Southern Africa • STAGE 1 Validation • 40 ETM+ acquisitions • 20 ETM+ scenes • 500,000 km2 • MODIS Burned area • = 0.85 “true” burned area Validation MCD45 Global Burned Area Product

  45. Burned Area Collection 5 Recommendations • This is a new product - there is no Collection 4 product • Product user manual and ATBD are posted on the MODIS Fire website http://modis-fire.umd.edu • Collection 5 production of MCDA1 has started • reprocessing of November 2000+ • forward production of February 2007+ • Note, changes to the Collection 5 algorithm are likely • Change descriptions will be posted on the MODIS Fire website • Product version reflected in the product PGEVersion metadata • Product quality issues posted on the LDOPE QA Web Site Recommendations MCD45 Global Burned Area Product

  46. THANKS MCD45 Global Burned Area Product

  47. Npass Nused Scheme of the detection Contextual iterative procedure: Burn candidates are selected if they provide persistent evidence of fire occurrence. The measured persistence varies depending on gaps in the reflectance time series and the timing of the fire relative to non-missing data, an iterative rather than simple thresholding approach is used Algorithm Background MCD45 Global Burned Area Product

More Related