A Step Towards census From satellite Koel Roychowdhury Simon - - PowerPoint PPT Presentation

A Step Towards census From satellite

Koel Roychowdhury Simon Jones Colin Arrowsmith Karin Reinke

School of Mathematical and Geospatial Sciences RMIT University, Melbourne, Australia

RMIT University Slide 2

Overview

• Introduction

– Research Objectives – Study Area – Datasets

• Methods
• Models
• Results
• Key Points
• Future Research

RMIT University Slide 3

INTRODUCTION

RMIT University Slide 4

Introduction

» DMSP-OLS night-time images primary source of data for project » DMSP-OLS used for a variety of applications (e.g. environmental sustainability, urban mapping and light pollution etc) » Problem of unavailability of census variables, particularly for small administrative regions » Propose an approach to produce surrogate census metrics at the sub-national level

RMIT University Slide 5

Research Objective

What is the utility of Average DN and radiance calibrated DMSP-OLS images for accurately predicting Indian census metrics at a sub- national level?

Achieve this by:

• Investigating statistical relationships between census metrics

and information derived from DMSP-OLS images

• Development of prediction models
• Validation and improvement of models
• Application of models to derive prediction maps of census

metrics

RMIT University Slide 6

Study Area

State of Maharashtra, India

• Size: > 300,000 km2
• Population: > 96 million
• Urban Population: 41

million in 378 urban centres

• Capital City: Mumbai

RMIT University Slide 7

Datasets Used

• Satellite Images

– DMSP-OLS

• Average Digital Number (DN) data (2001)
• Brightness data (2001)
• Census Data

– Primary Census Abstract (2001)

• Additional Census Data

– Maharashtra Development Report (2002)

RMIT University Slide 8

METHODS

RMIT University Slide 9

Census Vector Datasets Sampling of (24) Districts Statistical Testing Census Metrics (10 / 144) Selected Models Validation Final Outputs DMSP - OLS Images Mean and

• Std. Deviation

Brightness Annual Composite 2001 Average DN Annual Composite 2001 Intercalibration

Census data processing DMSP image processing Model development and implementation

Method Outline

RMIT University Slide 10

DMSP-OLS Image Processing

RMIT University Slide 11

DMSP-OLS Image: Intercalibration

• Differences in average DN between satellites
• Reference Image: captured by satellite F12 in 1999 over

Sicily

• Second order regression equation:

RMIT University Slide 12

DMSP-OLS Image: Results of Intercalibration

• For 2001, images obtained from satellites F14 and F15.

RMIT University Slide 13

DMSP-OLS Image: Selection of Average DN image

• F15 image has less

difference with F12 image after calibration

• F152001 image selected

for further analyses.

RMIT University Slide 14

DMSP-OLS Image: Mean and Standard Deviation

Summary Statistic District Highest mean Average DN Raigarh Lowest mean Average DN Gadchiroli Highest SD Average DN Nagpur Lowest SD Average DN Gadchiroli Highest mean brightness Nagpur and Pune (>20 watts/cm2/um) Lowest mean brightness Bhandara (<10 watts/cm2/um) Highest SD brightness Nagpur Lowest SD brightness Bhandara

RMIT University Slide 15

Census Data Processing

RMIT University Slide 16

Census Data: Sampling of Districts

• Random sampling.
• Process of random

number generation.

• 24 / 35 districts

selected randomly, 8 districts withheld for validation.

RMIT University Slide 17

Census Data: Statistical Tests

Statistical Tests Bootstrapping and correlation coefficients Tests for normal Distribution Histogram and Normal – Probability (N-P) Plots Skewness and Kurtosis

RMIT University Slide 18

Statistical Tests: Histogram and N-P plot

• Histogram distribution

and normal probability plot used to test for assumption of normality

• Example shown for

percentage of households with access to electricity

RMIT University Slide 19

Statistical Tests: Skewness and Kurtosis

• Kurtosis and skewness: measures of peakiness

and symmetry of the data.

• The further the values are from zero, the more

normal is the distribution.

• Tests to check whether zero is within 95%

confidence interval of skewness and kurtosis.

RMIT University Slide 20

Statistical Tests: Bootstrapping and Correlation

• Bootstrapping used to

help overcome problems

• f limited sample size
• 1000 bootstrap samples

created

• Correlation coefficients

(at 95% confidence)

• Examples shown for two

different census variables and average brightness

Bootstrap distribution of correlation coefficients between electricity and average brightness Bootstrap distribution of correlation coefficients between population density and average brightness

RMIT University Slide 21

• Demographic variables (e.g. population density)
• Economic variables (e.g. Per Capita District Domestic

Product)

• Social variables (e.g. percentage of households with

cars, jeeps and vans)

Selection of Census Variables: Common Census Metrics

RMIT University Slide 22

Final List of Census Metrics

• Number of households per square kilometre
• Total population density
• Urban population density
• Female literates per square kilometre
• Total number of workers per square kilometre.
• Percentage of households with car, jeep and van
• Percentage of households with access to electricity as

power source

• Percentage of households with television
• Percentage of permanent houses
• Per Capita District Domestic Product

RMIT University Slide 23

Models

RMIT University Slide 24

Models: Simple Linear Regression

• Models developed using single independent

variables obtained from DMSP-OLS images

• The independent variables used include:

– Mean brightness – Standard deviation of brightness – Mean average DN – Standard deviation of average DN

RMIT University Slide 25

Simple Regression Models

Urban population density per square Kilometre

Observed values Predicted Values a) Mean Brightness b) Standard Deviation Brightness c) Mean Average DN d) Standard Deviation Average DN r2 = 0.93, p<0.05 r2 = 0.79, p<0.05 r2 = 0.68, p<0.05 r2 = 0.42, p<0.05

RMIT University Slide 26

Validation: Simple Regression Models

25%

RMIT University Slide 27

Models: Multiple Linear Regression

Both mean and standard deviations considered together in each model 6 types of models examined:

– 2 models: Mean and standard deviation of brightness (with and without intercept) – 2 models: Mean and standard deviation of average DN (with and without intercept) – 2 models: all 4 independent variables (with and without intercept).

RMIT University Slide 28

Multiple Linear Regression Models with intercept

Observed values Predicted Values a) Mean and SD average DN b) Mean and SD Brightness c) Mean and SD average DN, and brightness r2 = 0.71, p<0.05 r2 = 0.92, p<0.05 r2 = 0.92, p<0.05

Urban population density per square Kilometre

RMIT University Slide 29

Observed values Predicted Values a) Mean and SD average DN b) Mean and SD Brightness c) Mean and SD average DN and brightness

Multiple Linear Regression Models without intercept

r2 = 0.68, p<0.05 r2 = 0.92, p<0.05 r2 = 0.90, p<0.05

Urban population density per square Kilometre

RMIT University Slide 30

Validation: Multiple Linear Regression Models

RMIT University Slide 31

Accepted Models

• Two models with the least errors were chosen at

the districts level

– Model with mean and SD of brightness – Model with all the four variables with intercept

• Sample equations:

) ( 5740684 . 15 12820 . 119 /

2

brightness mean Km ion banPopulat ExpectedUr × + − =

) ( 505353 . 4 ) ( 039980 . 22 ) ( 9849033 . 20 /

2

AverageDN SD DN average Mean brightness mean Km Population Urban Expected × + × − × = -108.45+16.93x(mean brightness)-2.48x(mean Average DN)

RMIT University Slide 32

Final Outputs

RMIT University Slide 33

Spatial Implementation of the Models

In addition to creating maps of predicted census metric values, maps showing the error between actual census data and the predicted data can also be generated

RMIT University Slide 34

Some other sample maps

(a) (b)

Summary

RMIT University Slide 36

Key points

• Annual composite of brightness image was better for prediction of

census metrics at sub-national level.

• Variables with absolute normal distribution over the districts (such as

sex ratio and education facilities per square Kilometre) do not have significant correlations with either brightness or average DN.

• Urban population density and PCDDP had higher correlations but

produced maximum errors in predicted values.

• Percentage of households with access to electricity had lower

correlation coefficients but was always predicted with least error.

RMIT University Slide 37

Future research

• Prediction of currently unavailable census variables for

administrative regions lower than districts.

• Assess the utility of DMSP-OLS images at the sub-district and

village (rural) level.

• Produce maps of the non-available metrics for finer spatial scales
• Delineate urban areas using both average DN and brightness

dataset and compare the results.

• Develop models for urban areas

RMIT University Slide 38

Acknowledgements

• My respected supervisors.
• AusAid for sponsoring the research
• Space Application Centre, ISRO, India for providing the

census datasets.

• My colleagues and friends in the Remote Sensing

Centre, School of Mathematics and Geospatial Sciences, RMIT University, Melbourne

Thank You