Building a Web Application with Open Source Technology
To Improve Access to Global Night-time Lights Data

Need for an Urbanization Science

Scholarship on cities is extensive, but our knowledge of urbanization is fragmented.



Cities are sources of environmental degradation.

Cities and the lifestyles they engender can be potential solutions to current and future environmental and sustainability challenges


Increasing amount of available information and data about cities that can help us address previous modeling barriers.

Lack a coherent understanding of the underlying urbanization processes that create urban places and interaction of these processes with other systems


The need for a holistic approach to urban development and human settlements and called for an integrated approach to planning and building sustainable cities and urban settlements to create livable places as a way to solve our local, regional, and even global environmental problems.



About the project

The proposed web application is about making nocturnal data more accessible for everyone.


  • Night-time lights data has the power to identify critical patterns
    and drive powerful insights into the spread of humans across the Earth.

  • Night-time lights data is too often inaccessible for people who need it.

  • We believe that by making this data more accessible
    and providing tools to make it more meaningful and actionable,
    more people become informed, engaged and empowered
    to address the issues that shape the future of our environment and our lives.



Modules

Intercalibrate

DMSP-OLS has no on-board calibration,
the individual composites were intercalibrated via an empirical procedure.
The objective of the intercalibration is to make it possible to compare the sum of lights index values from each year of the time series.

Store & Manage

Night-time satellite imagery will be stored and managed in an object-relational spatial database management system,
with the capabilities to conduct spatial query, spatial analysis, and statistical analysis.

Visualize

Map Visualization shows views of earth at night with interactive operations.
Chart and Table Visualization create meaningful information out of night-time lights data derived metrics.

Why is it a good idea?

Spatial ORDBMS

• Efficient storage as raster.
• Easy to manage.
• Easy to update.
• Automate imagery processing
and metrics generation procedures.

Visualizations

• Interactive map with query
and filtering capabilities.
• Compare night-time lights derived metrics
with other socio-economic indicators.
• Horizontal analysis
over a series of time periods.

Data Access

• Improve access to night-time lights data.
• Enable more applications
by providing access to night-time lights data
through developer's API.

Data

Global DMSP-OLS Nighttime Lights Time Series 1992 - 2013 (Version 4)


Work Flow


Store & Mange Night-time Imagery in PostgreSQL

Experiment with Raster Tile Size


  • DMSP-OLS is registered into 30arc sec grid. 120 x 120 pixel per tile = one degree per raster

  • Raster Size Raster Numbers Row Numbers Load Time Spatial Query Time
    1200 x 1200 36 x 14 504 56.34sec 1.2 hrs
    120 x 120 360 x 140 50,400 6 hrs 61 hrs
    60 x 60 720 x 280 201,600 73 hrs 87 hrs
  • Spatial SQL to Summarize and Aggregate Pixel Values to World Country Boundary

  • Computer Config:
    CPU: Intel Core i7-4770 3.4GHz
    RAM: 12.0 GB OS: 64-bit Windows 8.1 PostgreSQL 9.3 + PostGIS 2.1



Intercalibration

using empirically developed second-order regression functions


  • Samples of lighting from human settlements (cities and towns) were extracted from numerous candidate calibration areas and examined.

  • Data from satellite year F12-1999 had the highest digital values (DN = 63) because of saturation in the bright cores of urban centers and large gas flares. F121999 was used as the reference and the data from all other satellite years were adjusted to match the F121999 data range.

  • Unlit pixels in original image have nonzero DN because nonzero constants in quadratic polynomial regression equations were added to the pixels. A threshold is applied to only keep pixels with DN value > 3 in the final calibrated data.



Sum of Lights of U.S.

Before Intercalation


After Intercalation


Web Application


Future Research Direction




  • Integrate more night-time data source: VIIRS, ISS photos, NightSat.

  • Integrate environmental datasets (e.g. global NDVI)

  • Optimize database query functions, C implementation of PL/pgSQL.