1 / 13

Correction of bidirectionnal effects and impact on land cover classifications

This study explores the correction of bidirectional effects in land cover classifications and examines the impact of these corrections. The bidirectional compositing (BDC) approach is used to collect a constant number of data points to fit the bidirectional reflectance distribution function (BRDF), regardless of the acquisition date. The results demonstrate the importance of accurate cloud detection and highlight the challenges of noisy NDVI profiles in cloudy regions.

lastera
Download Presentation

Correction of bidirectionnal effects and impact on land cover classifications

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. Correction of bidirectionnal effects and impact on land cover classifications J-L. CHAMPEAUX METEO-FRANCE S. GARRIGUES METEO-FRANCE C. GOUVEIA ICAT, Universidade de Lisboa, Lisbon, Portugal EST, Instituto Politécnico de Setubal, Portugal P. BICHERON SCOT, Toulouse, France GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18-22 March 2002

  2. The usual strategy for directional correction: A 3 STEPS APPROACH NORMALISE the DATA qso; qvo=NADIR COLLECT RELIABLE CLOUD-FREE DATA FIT THE BRDF MODEL Roujean et al. (1992): MOD(s, v, ) =k0 +k1 f1 (s, v, ) + k2 f2 (s, v, ) Cloud mask For S Products?

  3. COLLECT THE N LAST CLOUD-FREE DATA FIT THE BRDF MODEL USE THE ADJUSTED BRDF TO NORMALISE THE CLEAR REFLECTANCES OF THE LAST 10 DAYS THE BIDIRECTIONAL COMPOSITING (BDC)P. MAISONGRANDE, B. DUCHEMIN, CESBIO MAIN IDEA : COLLECT A CONSTANT NUMBER OF DATA TO FIT THE BRDF, REGARDLESS TO THE DATE OF ACQUISITION AVERAGE OF THESE REFLECTANCES

  4. The BDC approach was applied here but S1 products are not optimal : - reduced data number and angular sampling - no possibility to track aerosol loading - cloud mask is not satisfactory SCOT CLOUD MASK: A pixel is cloudy if: B0(j) > 0.2 and B0(j) > 6* B0 Clim (month) And B0(j) > B2(j) METEO-France CLOUD MASK: A pixel is cloudy if: B0(j) > 0.11 And SWIR(j) > 0.09

  5. I IMPORTANCE OF CLOUD DETECTION  BDC  NDVI (SCOT mask) BDC NDVI (Meteo-France mask)

  6. Plot of 10-days NDVI profiles: MVC (red) , BDC (green)

  7. 10-days NDVI profiles for several methods MVC, BDC, Roujean model, daily data Blue: Roujean model Green:Roujean model Red: MVC Black: BDC

  8. CONCLUSION • The cloud mask is crucial • for use in kernel-driven BRDF models • (corrupted data distort the adjustment of • the regression) • After bidirectional corrections, • NDVI profiles are noisy mainly in cloudy regions • (due to the too small number of points used for the regression) • At this stage of the study, the landcover classifications made with data corrected from the bidirectionnal effects • do not improve the final results

More Related