Alexandre C. Coutinho Embrapa Monitoramento por Satélite PROCAM - USP BURNINGS OCCURENCY DYNAMCS IN THE STATE OF MATO GROSSO.

Alexandre C. Coutinho

Embrapa Monitoramento por SatélitePROCAM - USP

BURNINGS OCCURENCY DYNAMCS IN THE STATE OF

MATO GROSSO

• ACCURACY ESTIMATION OF BURNING DATA

• MUNICIPALITY BASED ANALYSIS

• MULTIVARIATED ANALYSIS - PRINCIPAL COMPONENTS

• FIRST COMPONENT CLASSIFICATION

• DYNAMICS CHARACTERIZATION

PROCEDURES

ACCURACY ESTIMATION OF BURNING DATA

1999

• 21 of 47 images

• spatial resolution - 60m

• geometrical resolution - 90m (1,5 pixel)

2000 / 2001

• 21 images

• spatial resolution - 30m

• geometrical resolution - 50m

1999

• 2899 burnings

• confidence level of 95%

• confifdence interval of 5%

2000 / 2001

• 716 burnings

• confidence level of 95%

• confifdence interval of 5%

SAMPLE

315 246

de de

339 250

NOAA - 12

ACCURACY ESTIMATION OF BURNING DATA

Stratified analysis based on distance

Identify geometric correction and commission errors

radius 1,5km; 3,0km and 6,0km around the coordinate

ACCURACY ESTIMATION OF BURNING DATA

1,5km

6,0km

3,0km

6,0km

ACCURACY ESTIMATION OF BURNING DATA

COMMISION ERROR ?

ACCURACY ESTIMATION OF BURNING DATA

ACCURACY ESTIMATION OF BURNING DATA

COMMISION ERROR ?

Samples:315 (1999)

Samples: 246 (2000/2001)

radius Km confirmed % %error confirmed % %error 1,5 202 64,1 35,9 150 61 39 3 263 83,5 16,5 194 78,9 21,1 6 301 95,6 4,4 226 91,9 8,1

commission 14 4,4 20 8,1

ACCURACY ESTIMATION OF BURNING DATA

COMMISSION ERROR ?

MUNICIPALITY BASED ANALYSIS

-1995-

MUNICIPALITY BASED ANALYSIS

-1996-

MUNICIPALITY BASED ANALYSIS

-1997-

MUNICIPALITY BASED ANALYSIS

-1998-

MUNICIPALITY BASED ANALYSIS

-1999-

MUNICIPALITY BASED ANALYSIS

-2000-

MUNICIPALITY BASED ANALYSIS

-2001-

MUNICIPALITY BASED ANALYSIS

-2002-

MUNICIPALITY BASED ANALYSIS

-2003-

MULTIVARIATED ANALYSIS - PRINCIPAL COMPONENTS

• data normalization

• Factor 1 explains 77,59%

Din. progressiva aceleradaInc. de qmd. extremamente alta

Din. progressiva moderadaInc. de qmd. alta

Din. estávelInc. de qmd. moderada

Din. estávelInc. de qmd. baixa

Din. Regressiva fracaInc. de qmd.muito baixa

LEGENDA

DYNAMICS CHARACTERIZATION