Presentación Índice de Sequía Agrícola VCI (Vegetation Condition Index)

Drought index based on remote sensing, using ’R’ and’raster’ package, in B́ıo-B́ıo region, Chile.

Francisco Zambrano Bigiarini

INIA

7 de diciembre de 2012

[email protected] (INIA) Drought index based en RS using R 7 de diciembre de 2012 1 / 35

Outline

1 General Review

2 Vegetation Condition Index (VCI)

3 Preliminary Results

4 Recommendations


General Review

Drought and its Classification

DefinitionThere is not a precise definition of drought

• ’drought means a sustained, extended deficiency in precipitation’ (WMO,1986)

• ’the percentage of years when crops fail from the lack of moisture’ (FAO, 1983)

• .....

Depending on the field/context, a drought event can be

1 Meteorological: drought is usually based on long-term precipitation departures fromnormal.

2 Hydrological: drought refers to deficiencies in surface and subsurface water supplies.

3 Agricultural: drought occurs when there is insufficient soil moisture to meet the needs of aparticular crop at a particular time.


General Review

Why use remote sensing for drought assessment?

Advantages

• Local and global monitoring.

• Spatial and temporal assesment.

• Widly used in the last two decades for monitoring vegetationcondition.

• A good choice for a developing country, which have insufficientdensity of weather stations

• Since 2000 we have data with highly temporal resolution and mediumspatial resolution from sensor MODIS.


General Review

Normalized Difference Vegetation Index (NDVI)

In short

It is a measure of the vegetation health.

NDV I =NIR− V IS

NIR+ V IS(1)

−1 ≤ NDV I ≤ 1

But in agricultural, its value is abovezero.

0 − > poor vegetation condition.1 − > good vegetation condition


General Review

Informatic tools for remote sensing and GIS

Wide variety of softwares

• Propietary (Very high cost applications): ENVI, ERDAS, ESRI, Matlab

• OpenSource: GRASS GIS, QGIS, ILWISS, R

What is needed?

• Procesing a huge amount of data raster.

• Perform sophisticated data analysis and graphics.

• Interact with other software.

• Work efficiently


Vegetation Condition Index (VCI)

VCI features (Kogan,1994)

• Derived from the NDVI

• Was designed to separate the weather related component of NDVIfrom ecological factors.

• Reflects the spatial and temporal vegetation variability.

• Its a scaling of the NDVI between its maximum and minimum.

• Has excellent ability to detect drought and to mesure time of itsonset, intensity, duration, and impact on vegetation.

• Its value goes between 0 and 100 %

V CIi = 100NDV Ii −NDV Imin

NDV Imax −NDV Imin(2)



Calculation example for a single pixel

1 We use data from MODIS sensor (2000-2012), 16 days composite.

2 Firstly we need an smoothed NDVI. (LOWESS (Cleveland,1981))









NDVI and VCI, 2000-2012


Preliminary Results

Whole region


Preliminary Results

Whole region


Preliminary Results

Spatial and temporal evolution in a drought season


Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Preliminary Results



Recommendations

Recommendations

• Analysis in the agricultural area (LANDCOVER).

• Consider the period most sensitive for crop growth.

• Validation of the VCI, make regression analysis with some variables.(crop yield, biomass, rainfall,...)

• Make research in the sense to extend the time series of NDVI images(NOAA, AVHHR).

• Sophisticated time series analysis.

• Extend the analysis to whole country.


Recommendations

Thanks?Questions?


Presentación Índice de Sequía Agrícola VCI (Vegetation Condition Index)

Documents

vegetation

temporal evolution

single pixel

preliminary

dv imin

remote sensing

general review

drought season