Loading...

Measuring economic activity from space  

| Updated: July 30, 2020 22:02:48


Measuring economic activity from space   

When flying into Dubai at night, Bangladeshi workers visiting the Middle East for the first time are astonished by the abundance of glittering lights on Dubai's buildings and structures, highways, and streets. Before setting foot on the ground, they know just how wealthy the Gulf emirate is. A 2014 night-shot of North Korea and South Korea, taken by a NASA astronaut, contrasts the dark, secluded hermit kingdom with the bright lights of one of Asia's affluent countries. The now-famous NASA picture tells better than a thousand words about the glaring (and embarrassing) disparity between  North and South Korea; it has now found a place in economic textbooks to educate students about the unconventional way of thinking about development. 'Seeing is believing' has been verified by further examples from around the world.

Numerous studies have found a strong correlation between night lights and economic activity across countries. It is easy to see why. Over the past decades, as millions of people in Asia escaped poverty, their houses too have lit up at night. We have seen how eagerly our village homes have gone from kerosene lamps to electric bulbs. The Economist puts it cogently: as close a tie as that between a person's height and hand size, night lights are linked to countries' GDP per person.

'Night lights' are the lights on the surface of the earth at night. They are captured with satellite imagery. This imagery indicates the intensity of light in tiny squares called pixels, which are one sq. m. A pixel indicates economic activity if its intensity exceeds a given threshold. The use of a threshold filter eliminates many lights such as small settlements, hilly areas, and flaring of burning natural gas that may not necessarily indicate economic activity. Night lights have become a popular proxy of economic activity among social scientists due to their timely availability, even in real-time for those who can afford it, and increasing accuracy. They provide information about a location at a more granular level.

Night-time lights offer a complementary way of measuring urbanisation in a country. The conventional approach of population density and revenue-based classification of a municipality is still useful, but as already stated, it is a prolonged process. Based on a weighted average of A, B, and C categorisation of municipalities by the LGED, we have come up with a threshold point and designated upazilas as 'urban' whose intensity of lights visible at night exceeds this threshold.

In recent research, at the Centre of Urban Studies and Sustainable Development at East West University using night light data across 544 upazilas we explore the extent of economic activity in these administrative units over the years 1992-2013. The current practice of relying on census and household income and expenditure survey to know the financial condition of households has a time lag of 5-10 years. Whereas, using night lights data, we can get a proxy of the economic health of upazilas on a daily, weekly, or monthly basis. As of now, our use is limited by the data that NOAAA generously provides, but this is enough to show us its future potential.

The figure below shows the growth of night light intensity for the top and bottom five upazilas over the period 1992-2013. Tungi Para in Gopalganj district ranks as the most economically active upazila in Bangladesh. Particularly in the years 1998-2000 and 2008-2013, Tungi Para saw the strongest night lights luminosity. Next comes Bhaluka in Mymensingh district, where night lights have steadily increased since 1992. Located a short distance from Dhaka, garment factories, retail showrooms, and manufacturing plants (e.g., Coca Cola, Butterfly-LG) are housed in the Bhaluka upazila.

Using the threshold of the night-time light, between 1992 and 2013, the number of urban areas increased from 126 to 202, a 60 per cent increase in 22 years. Most sadar upazilas have become urban areas, according to the night-time lights satellite imagery. However, the bulk of the urban areas are concentrated in Dhaka, Chittagong and Sylhet divisions, suggesting that the remaining four divisions have fallen behind on economic activity. This result echoes the findings of numerous studies on regional economic disparity in Bangladesh.

In the Dhaka district, Gazipur sadar upazila has registered the highest intensity of night lights. Over time, many residential and commercials units have moved to Gazipur, which became part of greater Dhaka. What is more interesting is that several upazilas of B and C categories exhibit the level of night lights intensity comparable to category A. These are Kasba, Raozan, Bancharampur, and Zakiganj, to cite a few examples. Conversely, several A category municipals have seen a decline in night lights intensity, implying that these areas have lost economic dynamism over time. These include Amtali, Bakerganj, and Mujib Nagar, among several others.

Miyabazar, a town Chauddagram upazila in Comilla district, is a curious story. Until only a few years ago, Miyabazar was practically an open field with a few scattered houses. But once the highway was built, hotels kept coming. Over the period 1992-2013, night-time lights in Chauddagram grew at an annual rate of 9.2 per cent, helping Miyabazar to become a 24-7 living community. Over the years, several ready-made garment factories also relocated to Miyabazar to take advantage of its transportation routes, among other attributes. One example is that of Kuddus Garments factory, helping to cut congestion and avoid high rents in Dhaka.

To sum up, night time lights data enables real-time monitoring of urban development patterns over time. Defining a metropolitan area (or, for that matter, a city) is no longer limited to administrative boundaries as businesses and economic linkages among people do not stop at union, upazila, or district borders. As cities are essentially integrated labour markets, statistical agencies in developed countries define metropolitan areas on the basis of commuting flows of the workforce. Absent such information, researchers are increasingly relying on night-time satellite images to identify urban areas in developing countries such as Brazil and India.

Sadly, though, for such a productive and informative indicator, we have to rely on foreign expertise, which causes unnecessary delay. The Bangladesh Space Research and Remote Sensing Organisation (SPARRSO) can be delegated to generate such data on a real-time basis, which could also help policymakers and businesses gauge real economic activity before the official data is published. For example, in the March 2020 lockdown, night light radiance in India's three big cities (Bengaluru, Delhi, and Mumbai) fell over 30 per cent, compared with the same period last year. In March 2020, when China eased the lockdown, night lights in industrial parks across the country also lit up, though their levels remained below potential.

Besides mapping global poverty and population density, night lights are being used to track border conflicts, understand community resilience, fishing activity, and to study how childhood and adolescence exposure to night lights affect brain development and human behaviour. Bangladesh economic policy has yet to utilise the full potential of this gift from outer space.

 

The authors are, respectively, research associate, professor of economics, and university professor at East West University.

[email protected]

Share if you like

Filter By Topic