DROUGHT ASSESSMENT USING SATELLITE

DERIVED METEOROLOGICAL PARAMETERS AND

NDVI IN POTOHAR REGION

Researcher: Saad-ul-HaqueSupervisor: Dr. Badar Ghauri

Department of RS & GIScInstitute of Space Technology

Karachi Campus

Overview

What is Drought Historical perspective Study Area Objectives Problem Statement Dataset Methodology Results Significance of study

2

What is Drought?

A natural hazard and is a part of climatic condition for all regions of the world Condition of moisture deficit caused by a certain climatic conditions occurring at a

specific location for a specific duration Stems from the lack of precipitation, precipitation deficiency for a season, a year or

longer Triggered, when water supplies become insufficient to meet the requirements

3

Pakistan in Historical Perspective

During 1999-2000- 2.21 million people population

affected- 23.51 million livestock

A weak drought developed in mid-2009 to mid-2010

- Resulting 30% below normal monsoon rains in northern area.

- 2.5% drop in crop yield (Wheat)

Pakistan Metrological Department & Agricultural Statistics of Pakistan 2010-114

Drought Stages

Wilhite, D.A.; and M.H. Glantz. 1985. Understanding the Drought Phenomenon: The Role of Definitions. Water International 10(3):111120

5

Study Area

The Barani (rainfed) Land covers the Salt Range and the Pothohar Plateau

The climate is semi-arid and sub tropical

There is no means of canal irrigation system in this area and the agriculture primarily depends on the amount of rainfall

Being hilly in nature the water retention is limited due to peculiar drainage pattern of Potohar region

In the past, several droughts have experienced in this area that left a drastic agricultural and socioeconomic impact

6

Problem Statement

- The study area is more vulnerable to short term droughts especially in winters due to the absence of Western Depression

- Anomaly of temperatures, humidity and increasing solar flux also accelerate the reduction of water content in soil and atmosphere

- Occurrence of such conditions and prolonged absence of regular rainfall lead to drought conditions, affecting crop yields and thus economy of the country

Objectives

Utilization of Geo-spatial techniques (RS & GIS) for Drought monitoring, as a cost effective and precise option compared to traditional approaches

Identification and assessment of drought severity and anomalies in the Potohar region during 2009-2010

Delineation of significant meteorological parameters that may contribute to the occurrence of short term meteorological droughts.

Development of spatial extent of drought severity map that can help to identify the most affected area in a region.

7

Dataset used:

8

MODIS NDVI Product

Moderate Resolution Imaging Spectroradiometer (MODIS) Terra Normalized Difference Vegetation Index (NDVI) product (MOD13Q1)

The dataset having resolution of 250 meters The data has 16 days temporal resolution Total 303 images were obtained for the period from Feb 2000 to Mar 2013

NCEP Meteorological Data

Satellite data from National Centers for Environmental Prediction (NCEP) Climate Forecast System Reanalysis (CFSR).

The global atmospheric resolution is approximately 38 km, extending to 0.5beyond the tropics.

Meteorological data includes:o Precipitation o Maximum Temperature, o Minimum Temperature, o Humidity, o Solar Radiation

9

Methodology

10

11

Identification of Drought period

The multi temporal images in NDVI for year 2002 to 2013 were stacked into a single imagery.

The 13 year collective NDVI image was used to obtain temporal signature of minimum, maximum, and mean NDVI values.

The signatures showing NDVI trend and deviation from mean NDVI was plotted.

12

0.10.20.30.40.50.60.70.80.9

1Time Series Analysis of NDVI

MAX MEAN

13

00.10.20.30.40.5

MEAN NDVI (Winter Season )

MEAN NDVI MEAN

14

15

Vegetation Condition Index (VCI)

The VCI is a pixel-wise normalization of NDVI that is useful for making relative assessments of changes in the NDVI signal by filtering out the contribution of local geographic resources to the spatial variability of NDVI.

It shows how close the NDVI of the current month is to the minimum NDVI calculated from the long-term record.

VCI j = (NDVI j NDVI min) * 100(NDVI max NDVI min)

Where, - NDVI max and NDVI min are calculated from the long-term record (e.g., 23 years) for that month (or week)

The vegetation condition/health represented by VCI is measurement that is represented in percentage.

16

VCI Table

Value in Percentage Vegetation Condition

Between 100 50 % Optimal / Above normal

Around 50 % Fair

Below 50 % Drought

Below 35 % Severe Drought

Kogan., F. N. (2000). Contribution of Remote Sensing to Drought Early Warning. NOAA, NESDIS17

14 Sep 09

18

NCEP DATA POINTS

19

14 Sep 09

20

14 Sep 09

21

14 Sep 09

22

23

Standardized Precipitation Index

The SPI is an expression that calculates and represents the deficiency of rainfall for different time scales.

Six-month based SPI indicates medium term trend in precipitation pattern.

Therefore six month SPI was calculated for the winter drought season for the year 2009 2010, i.e., cultivation period September-February.

Z score = (X-) / Standard DeviationWhere,

- Z score is SPI value- X is precipitation value- represents mean of six month rainfall

24

SPI Table

Description Standardized precipitationIndex (SPI)

No drought -0.5 and aboveAbnormally Dry -0.5 - -0.7Moderate Drought -0.8 - -1.2Severe Drought -1.3 - -1.5Extreme Drought -1.6 - -1.9Very Severe Drought -2 or less

Source: (U.S. National Drought Mitigation Centre)

25

14 Sep 09

26

NDVI based Correlation Trend

27

Chart1

0.2277-0.29566-0.339340.148220.07631

0.1747-0.29612-0.349280.191630.04115

0.12081-0.24076-0.279820.15312-0.20587

0.32808-0.24814-0.318280.014620.27874

0.30596-0.20863-0.25417-0.09930.30551

0.01717-0.0744-0.117060.013630.19445

0.03079-0.10298-0.13659-0.174110.24937

0.0822-0.11633-0.116910.008250.18048

0.16251-0.18625-0.1706-0.008690.24347

0.24458-0.27452-0.30019-0.135020.2604

Humidity

Max Temp

Min Temp

Solar

SPI

NDVI base Correlation Trend with Meterological Parameters

Sheet1

S.NoHumidityMax TempMin TempSolarWindSPI

Sep-090.2277-0.29566-0.339340.148220.333110.07631

0.1747-0.29612-0.349280.191630.35020.04115

Oct-090.12081-0.24076-0.279820.153120.32707-0.20587

0.32808-0.24814-0.318280.014620.33990.27874

Nov-090.30596-0.20863-0.25417-0.09930.291460.30551

0.01717-0.0744-0.117060.013630.231080.19445

Dec-090.03079-0.10298-0.13659-0.174110.178410.24937

0.0822-0.11633-0.116910.008250.166850.18048

Jan-100.16251-0.18625-0.1706-0.008690.154760.24347

Feb-100.24458-0.27452-0.30019-0.135020.051030.2604

Sheet1

Humidity

Max Temp

Min Temp

Solar

Wind

SPI

NDVI base Correlation Trend with Meterological Parameters

Fuzzy Overlay

The Fuzzy Overlay tool allows analysis of possibility of a phenomenon belonging tomultiple sets in a multi-criteria overlay analysis.

Fuzzy Overlay not only determines what sets the phenomenon is possibly a member of,it also analyzes the relationships between the memberships of the multiple sets.

Overlay type lists the methods available to combine the data based on set theoryanalysis.

The available methods are Fuzzy And, Fuzzy Or, Fuzzy Product, Fuzzy Sum, and FuzzyGamma. Each approach provides a different aspect of each cell's membership to themultiple input criteria.

28

Discretization criteria for Fuzzy overlay and VCI

Fuzzy Overlay Category VCI values Category

-3 to -1.5 Extremely dry 0 to 34 Very Low

-1.49 to -1.0 Dry 35 to 45 Low

-0.99 to 0.99 Normal 46 to 65 Normal

1 to 1.5 Wet 66 to 85 High

1.5 to 3 Extremely Wet 86 to 100 Very High

Spatial Data Mining for Drought Monitoring: An Approach Using temporal NDVI and Rainfall RelationshipINTERNATIONAL INSTITUTE FOR GEO-INFORMATION SCIENCE AND EARTH OBSERVATION ENSCHEDE, THE NETHERLANDS

29

Overall Spatial Extent of Meteorological Drought

Drought severity map was generated based on the discretization criteria for resultedfuzzy overlay and VCI into 5 different categories.

Each 16 day Composite image from September 2009 to February 2010 were thanclassified using raster reclassify tool, into following classes:

Very Severe Drought Severe Drought Moderate Drought Least Drought No Drought affected areas

30

14 Sep 09

Overall Spatial Extent of Meteorological Drought

No Drought

Least Drought

Moderate

Drought

Severe Drought

Very Severe

Drought

31

Results

There was an unexpected decrease in NDVI value in 2009-2010, which caused a short termdrought.

A negative relationship between NDVI and SPI, in month of October was observed thattriggered the condition of drought for that period.

A decreasing trend was observed in relationship between NDVI and humidity. This could beuseful to identify water stresses on crops.

Overall severity map generated gives a significant understanding about the areas undervarious degrees of drought stress

32

Significance of the study

Nearly 62 percent of the population in rural areas is directly or indirectly associated with

agriculture sector for their livelihood, such studies would have direct impact on their

socio-economic condition

This study has the potential to act as a benchmark for understanding drought

phenomenon using satellite-based meteorological data keeping in view the fact that

other means of irrigation were not available in the study area

This study provided a methodology which can be used in future to identify any drought

condition occurring in this region. Thus it could be used as an early warning tool for

farmers, food availability, crop insurance agencies, water management and policy

makers.

33

34

DROUGHT ASSESSMENT USING SATELLITE DERIVED METEOROLOGICAL PARAMETERS AND NDVI IN POTOHAR REGIONOverviewWhat is Drought?Pakistan in Historical PerspectiveDrought StagesSlide Number 6Problem StatementDataset used:MODIS NDVI ProductMethodologySlide Number 11Identification of Drought periodSlide Number 13Slide Number 14Slide Number 15Vegetation Condition Index (VCI)VCI TableSlide Number 18Slide Number 19Slide Number 20Slide Number 21Slide Number 22Slide Number 23Standardized Precipitation IndexSPI TableSlide Number 26NDVI based Correlation TrendFuzzy OverlayDiscretization criteria for Fuzzy overlay and VCI Overall Spatial Extent of Meteorological DroughtSlide Number 31ResultsSignificance of the studySlide Number 34