GLC 2000 land cover map of the desert

  • Slides: 14
Download presentation
GLC 2000 land cover map of the desert locust recession area First results E.

GLC 2000 land cover map of the desert locust recession area First results E. Bartholomé, JRC Ispra M. Cherlet, FAO Rome 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

The desert locust recession area 19°W – 80° 30’E, 46°N – 4°N Main focus

The desert locust recession area 19°W – 80° 30’E, 46°N – 4°N Main focus on drylands 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Input data • S 10 NDVI for 2000 • Re-scaled in vegetation cover percentage

Input data • S 10 NDVI for 2000 • Re-scaled in vegetation cover percentage • Soil effect removed (per-pixel method) • Temporal profile smoothed (see ad hoc presentation) 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Classification procedure Based on temporal profiles Rigid tresholds on: • Max vegetation cover in

Classification procedure Based on temporal profiles Rigid tresholds on: • Max vegetation cover in 2000 • Season duration cover • Seasonality: max in summer or winter, or double seasonality 1/3 time 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Rationale • Cover max – corresponds to max development of vegetation – Best correlates

Rationale • Cover max – corresponds to max development of vegetation – Best correlates with above-ground biomass • duration Duration close to phenological definition, but time lag 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Classes Max cover 1 BARE AREAS 2 SPARSE HERBACEOUS VEGETATION - NOT ANNUAL TBC

Classes Max cover 1 BARE AREAS 2 SPARSE HERBACEOUS VEGETATION - NOT ANNUAL TBC 100% 3 SPARSE HERBACEOUS VEG 9 4 CLOSED-OPEN HERBACEOUS VEG 5 SPARSE DECIDUOUS SHRUB AND SPARSE HERBACEOUS VEGETATION 75% 6 OPEN DECIDUOUS SHRUB AND OPEN HERBACEOUS VEGETATION 6 50% 8 7 SPARSE DECIDUOUS SHRUB AND SPARSE HERBACEOUS VEGETATION AND SPARSE TREE LAYER 4 10 6 8 5 7 8 OPEN DECIDUOUS SHRUB AND OPEN HERBACEOUS VEGETATION AND SPARSE TREE LAYER 9 CLOSED-OPEN DECIDUOUS SHRUB AND CLOSED- OPEN HERBACEOUS VEGETATION AND OPEN TREE LAYER 10 NOT DEFINED 25% 10% 5% 1 6/7/2021 2 3 3 6 9 12 months 1 The same applies for winter vegetation Growth and double season growth GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002 duration

Result 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March

Result 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Close-up (1) Sarir (Libya) Irrigated agriculture 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC

Close-up (1) Sarir (Libya) Irrigated agriculture 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Close-up (2) Luxor (Egypt) double growing season 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP

Close-up (2) Luxor (Egypt) double growing season 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Close-up (3) Senegal - Mauritania 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC –

Close-up (3) Senegal - Mauritania 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Observation 1 • Noise in the 3 -5% cover range Haruj es Sawda Volcano

Observation 1 • Noise in the 3 -5% cover range Haruj es Sawda Volcano (Libya) 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Observation 2 • Specific year 2000 situation 1999 -2000 Land cover map 6/7/2021 2000

Observation 2 • Specific year 2000 situation 1999 -2000 Land cover map 6/7/2021 2000 -2001 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

Observation 3 • Cropland not detected per se: low % of surface 6/7/2021 GLC

Observation 3 • Cropland not detected per se: low % of surface 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002

conclusions • The method seems efficient for dryland vegetation (bare soil correctly mapped!) •

conclusions • The method seems efficient for dryland vegetation (bare soil correctly mapped!) • The method is sensitive to the specific year for very sparse vegetation • Some “soil contamination” • Need for regional analysis to label the land cover classes 6/7/2021 GLC 2000 – “FIRST RESULTS” WORKSHOP JRC – Ispra, 18 -22 March 2002