Satellite imagery gauges economic growth and land use change

I got an email tip on this article from which has some relevance to the work done by the project in that it shows clearly the impact of urbanization. While Hansen et al (GISS) uses “nightlights” in the USA to gauge “urbanness” of a station’s surroundings, they only use one source image from the Defense Meteorological Satellite Program from 1995.  You can do the same yourself in Google Earth. Clearly from this example, GISS should be updating that source image if they are to get anything remotely representative of a true measure of urbanization around a climate station

This image is also illustrative to what Dr. Pielke Sr. and others have been saying for sometime. Impacts of Land use and land cover change can affect the surface temperature measurement environment over time and should be considered in any assesment of local/regional climate trends.

Of course the most striking evidence of economic growth measured by nightlights comes from this DMSP image of North and South Korea:


h/t to WUWT reader and climate blogger Warren Meyer for the link. – Anthony

From Reposted per their terms of use.

Measuring economic growth from outer space

J. Vernon Henderson Adam Storeygard David N. Weil 2 September 2009

GDP data is often poorly measured, especially for sub-Saharan Africa. This column shows that satellite data on lights at night can be used to enhance the quality of GDP growth measures. Using rainfall and satellite data, it also shows that growth of immediate agricultural hinterland of a sub-Saharan city spurs growth of the city.

GDP growth is poorly measured for many countries (Johnson, Larson, Papageorgiou, and Subramanian, 2009) and rarely measured for cities at all. The Penn World Tables rank countries by the quality of their GDP and price data, with grades A-D. Almost all sub-Saharan African countries get a grade of C or D, to be interpreted roughly as a 30% or 40% margin of error (Deaton and Heston, 2008). Given the low quality of GDP measures for countries and the almost total absence of GDP measures for sub-national units such as cities, we propose a readily available proxy: satellite data on lights at night. The best use of lights data is to examine growth in GDP rather than GDP levels, so that cross-country differences in how lights spatially and culturally reflect consumption are differenced out.

We start by examining cross-country GDP growth rates, focusing on the period 1992-2003, and develop a statistical framework for optimally combining the growth in lights measure for each country with estimates of GDP growth from the World Development Indicators. We first establish that changes in lights are well related to particular positive or negative economic growth episodes for particular regions and times and, more generally, that growth in lights is a good predictor of growth in GDP measures. As an illustration (Elvidge et al, 2005), Figure 1 contrasts the big increase in lights from 1992 to 2002 in the Eastern European countries of Poland, Hungary, and Romania with the distinct dimming of lights to the east in the former Soviet Republics of Moldova and the Ukraine, which endured a harsh transition process.

Figure 1. Eastern Europe in lights

Click for a larger image

Next, we develop a framework to optimally combine measured GDP growth with growth in lights to obtain a best estimate of true GDP growth. The objective is to minimise the variance of true GDP growth from its best estimate. The weights placed on the World Bank GDP growth measure and the lights growth measure depend in part on the ratio of signal to total variance in the World Bank measure.

Applying our method to the countries given a data quality grade D in the Penn World Tables, we get estimates of true GDP growth that are starkly different from conventional measures. We assume the World Bank measures have a signal to total variance ratio of 0.75. This is likely to be conservative since grade D countries are expected to GDP levels measured with a 40% margin of error. As examples of the application, for the Democratic Republic of Congo, lights suggest a 2.4% annual growth rate in GDP, while official estimates suggest a -2.6% growth over the same time period. The Congo seems to be growing a lot faster than official estimates suggest. At the other extreme, Myanmar has an official growth rate of 8.6% a year, but the lights data imply only a 3.4% annual growth rate. Combining the two measures using the hypothesised signal to total variance ratio, the true growth rate estimates for Congo and Myanmar are 0.08% and 4.6% for 1992-2003.

Finally, we turn to a long standing debate in developing countries about whether growth of the immediate agricultural hinterland of a city spurs growth of the city. We use annual rainfall data for the hinterlands of 541 cities in sub-Saharan Africa as our exogenous source of agricultural growth. We find that increases in rainfall have big positive effects on city growth as measured by changes in night lights, confirming the casual impression that African cities and towns are heavily dependent on the economic health of their immediate hinterlands. Lights in a given year are affected not just by rain in the same year but also by rain in the previous two or even three years. Not surprisingly the effects are smaller for the primate cities of a country which are less dependent on their agricultural hinterlands. But overall city growth is closely connected to local hinterland growth.


Deaton, Angus and Alan Heston. 2008. “Understanding PPPs and PPP-based National Accounts.” NBER Working Paper 14499.

Elvidge, Christopher D., Kimberley E. Baugh, Jeffrey M. Safran, Benjamin T. Tuttle, Ara T. Howard, Patrick J. Hayes, and Edward H. Erwin. 2005. “Preliminary Results From Nighttime Lights Change Detection.” International Archives of Photogrammetry, Remote Sensing and Spatial Information Sciences, 36(8).

Henderson, J. Vernon, Adam Storeygard, and David N. Weil (2009). “Measuring Economic Growth from Outer Space.” NBER Working Paper 15199.

Johnson, Simon, William Larson, Chris Papageorgiou, and Arvind Subramanian. 2009. “Is Newer Better? The Penn World Table Revisions and the Cross-Country Growth Literature.“ Unpublished.

0 0 votes
Article Rating
Newest Most Voted
Inline Feedbacks
View all comments
September 3, 2009 9:50 am

Food and energy

Gordon Ford
September 3, 2009 9:55 am

Would not night time lights be a reasonable proxy for the calculating the change in the Urban Heat Island effect?

September 3, 2009 10:02 am

How precise can such measurement be?
The lamps could have been changed to brighter ones. The surface reflecting the light could have changed, because usually street lights have reflectors on top to force the light down and reduce nightsky light pollution. People might put more than one light around their hourse… never seen those garden lights?
Also, where there are lights does not necessarely mean that there are houses and people.
I think a day picture would really show clearly if there is a change in urbanization.

September 3, 2009 10:10 am

North Korea looks like the best example of what the warmers would like the rest of the world’s economies to look like. There’s not a lot of CO2 being produced at night in that country to keep the lights on.
Perhaps we could cut CO2 emissions in the developed world by just passing a law that would require blackouts similar to WWII at 10pm every night. Who needs all those streetlights on anyway when everyone is at home blogging, watching cable TV or Youtube.

John Laidlaw
September 3, 2009 10:20 am

It’s a pretty good proxy – not 100%, but a good indicator. But here’s a wild and wacky thought; same scale, same resolution (or close to, given the longer wavelength) infrared images showing the amount of heat being radiated. Especially if it were possible to compare it to non-urbanised areas close by. Just a thought.
I do like William’s idea of a blackout. I miss seeing the Milky Way… and think of the energy that would save.

John Galt
September 3, 2009 10:21 am

Isn’t GISS part of NASA? I wonder where GISS could get some up-to-date satellite images? Hmmm….

John S.
September 3, 2009 10:38 am

North Korea, or eastern China, is one region in which I’d like to find a century-long temperature record from a site that shows relatively few night lights. Alas, not even Pyonyang has a record of sufficient length, and the smaller towns in China are not being updated by GHCN. This leaves only records from the burgeoning mega-cities to provide regional indications of recent temparatures. I was hoping that, on his recent trip to China, Anthony might have established contacts there that could provide the current data that GHCN has suppressed. I’m sure that many of us would be interested in seeing what the less-UHI-affected smaller towns show. Can you help us, Anthony?

September 3, 2009 10:41 am

Yes, William, I also was struck by the differences between the relatively free South Korea and the totalitarian North Korea — and how many of the more rabid warmists would have us emulate the emissions (and therefore the economic conditions) of the North rather than the South.
I’m sorry, but CO2 emissions go hand in hand with wealth and will for some time to come. There is no immediate way of significantly reducing CO2 without reducing wealth.

Bob H.
September 3, 2009 11:03 am

A couple of points. I have done roadway lighting design, and can safely say that the lighting, particularly in major cities has been upgraded over the last 50 or so years. It started out with Mercury Vapor, transitioned to High Pressure Sodium, and is now transitioning to Metal Halide and LED. Generally speaking, the lumen output of the lighting has not changed much for an individual lamp, but the wattage has decreased over time. The typical street corner light uses a 400 Watt High Pressure Sodium bulb emitting 50,000 lumens. High mast lighting can have upwards of 1,000,000 lumens per mast at a consumption of 6,000 to 8,000 watts.
While it is true that cutoff lenses have replaced more difuse lenses, the output of the bulb stays the same. The light seen from space, in most cases is light reflected from the ground rather than radiated. The exceptions would be billboards, highway signs,flagpole lights, and a few others, but which don’t make a large contribution to the light observed from space.

John Egan
September 3, 2009 11:03 am

There is an interesting study which might be done based upon these two images.
Whereas Poland, Hungary, and Romania has more light – especially Poland –
Moldava and Ukraine had less – especially Moldova.
What if someone were to look at temperatures in Western Galicia (Poland), Eastern Galicia (Ukraine), Moldavia (Romania), and Moldova for this time period and see if there is any correlation?
There might be some good info out there.

September 3, 2009 11:10 am

William (10:10:06) :
“Perhaps we could cut CO2 emissions in the developed world by just passing a law that would require blackouts similar to WWII at 10pm every night. ”
That’s a great Idea. The USA would be black just like north korea, with the sole spot of light being Al Gore’s house!

September 3, 2009 11:48 am

John S. (10:38:18) :
Why North Korea, or eastern China temperature records would be even remotely interesting? If you want to find some stations with reliable records here they are:
They are remarkably flat, aren’t they? I explain it due to virtually no noise due to daytime/nighttime temperature differences. Ironically, according to warmists it is the poles, that are expected to warm the most!

Frederick Michael
September 3, 2009 11:55 am

“Team America, World Police” was right. North Korea sure looks like a lonely place.

September 3, 2009 12:16 pm

And all those new people in those newly occupied areas are taking water that would normally be sitting in some lake or the ocean and blowing in out on their property, thereby increasing water vapor in the atmos. The spread of farms would make the biggest impact in this direction.

September 3, 2009 12:29 pm

There are numerous studies linking the urban heat island effect (rising temperatures) with population growth, here is one.
Al Gore’s filmmaker foes branded ‘Hitler’s henchmen’ by environmentalists,

September 3, 2009 12:36 pm

John S. wrote: “North Korea, or eastern China, is one region in which I’d like to find a century-long temperature record from a site that shows relatively few night lights.”
Ren et al. used 282 40-year temperature records (1961-2000) from northern China and broke them into groups (rural, small city, medium city, large city and metropolis). As you might expect, the larger the city the higher the temperature trend recorded.
Here are two descriptions of the work:

September 3, 2009 1:10 pm

I am distrustful of these displays for the reasons Ray mentions and others. The Sierra Club for while handed out postcards of Montana with the road system on it. It looked like 40% of the state was paved over! But of course it was just standard propaganda using different scales for different items on the same projection. The roads would have to have been better than 1/4 mile wide to show up as they did. The purpose of course was just the general propaganda push that it was nearly too late to save something or other (I lost track of just what now).
I’ve also seen these “ights from space”projections before. The trouble is, when I fly at night I notice that it doesn’t look like that at all. There are vast areas, even in relatively populated areas which are dark. So I wonder what the ‘digital scale’ in this one really is displaying and I don’t really pay a lot of attention to it.

September 3, 2009 1:15 pm

I don’t know if this is related, but when scientists start blathering about temps, ice ages, and getting the start date of the industrial revolution wrong, I shake my head and turn here.
Anthony: Perhaps you can read this and shed light in a new post? Or comment here? Thanks.

September 3, 2009 1:16 pm

Bob H. (11:03:06) :
The fact that the technology has been changing, this study is then invalidated. You can bet that the poor countries are not using state-of-the-art lamps. LED lamps will almost not be registered with this satellite since they can be considered “cold”.
So, what’s the point of all this if not to be asking for more big grants.
CO2 is not a pollutant… political hot air is.

Mike O
September 3, 2009 1:25 pm

My son tells me that North Korea enforces a nightly blackout to prevent the imperiList dogs from finding targets for their imminent bombing raids.

Mike O
September 3, 2009 1:27 pm

Imperialist … Stupid iPhone, thinks it knows more than I!

Frank Lansner
September 3, 2009 1:36 pm

OT: A little news from Copenhagen, the climate meeting:
The Danish authorities has reserved hotel rooms for more approx 50 mio US dollars.
But… By now it actually seems that only 20% of the climate-visitors appears to be coming afterall…
(Leaving the authorities with a gigantic loss of money for the reservations..!)

September 3, 2009 2:03 pm

Gordon Ford :

Would not night time lights be a reasonable proxy for the calculating the change in the Urban Heat Island effect?

This is something that can be quantitatively established, and that I would like to see tested with the Surface Stations database.

Greg Cavanagh
September 3, 2009 2:34 pm

Ray (13:16:47) :
LED lamps will almost not be registered with this satellite since they can be considered “cold”.
That would depend on what the salellite was recording, heat or visible light.
There is no reason why poorer countries can’t or don’t use LED lights, as they cost the same to install. Bob’s point was that lighting technology has changed in the last 30 years from Murcury to Sodium to LED, but the cost to install each is the same. (I’m a road designer not a lighting designer, but I get to authorise the bills).

September 3, 2009 2:40 pm

Hansen uses these in the US adjustments, doesn’t he? Might explain why the US temp series from GISS warms less than NCDC.

September 3, 2009 2:58 pm

Oh. Right. Up top. Sorry.

September 3, 2009 3:01 pm

Some posters apparently think GDP is a temperature measurement. Thus they assume that the paper under discussion must be a warmist willie propaganda vehicle.
Okay, fellers, nobody tole you this, but this here GDP is GROSS DOMESTIC PRODUCT!
Measuring GDP is harder than hell in some countries, especially smaller ones with little financial data. The paper simply espouses using nighttime light changes as a proxy for GDP changes in those areas. This hypothesis is correlated by comparing light changes to GDP growth in countries with sufficient economic information to provide valid data.
See the paper for a complete explanation. The paper is probably excellent. Is their method precise? Hell no, nor does it pretend to be. The authors are trying to improve the GDP data, not nail it down to three sig figs. I think they’ve done a fine job.

September 3, 2009 3:29 pm

Frank Lansner (13:36:13) : said
“OT: A little news from Copenhagen, the climate meeting:
The Danish authorities has reserved hotel rooms for more approx 50 mio US dollars.
But… By now it actually seems that only 20% of the climate-visitors appears to be coming afterall…
(Leaving the authorities with a gigantic loss of money for the reservations..!)”
Can you make a block booking of 250 rooms at a very good price and get 250 delegate tickets?
Now, who fancies a trip to Copenhagen as part of the WUWT delegation?

September 3, 2009 3:30 pm

Seems to me that this sort of thing would be right up Google Earth’s alley for near real time overlays. Sure would be interesting if someone who knew how, would create a downloadable overlay/app we could all access. Might have some national security implications tho ( for various countries), come to think of it.

September 3, 2009 3:55 pm

john S,
An example of rapid urbanization is Singapore, which shows a pronounced upward trend in temperatures, 0.5C in the last 25 years, more than double the global average over that period.
I believe one of the weather stations is at Changi Airport which has grown enormously over the last 25 years , and may show a particulary large UHI.

September 3, 2009 4:21 pm

The proxy of a proxy is still a proxy.
With such proxy, maybe we should start living underground… just to mess up with it.

John Trigge
September 3, 2009 4:21 pm

I should not be surprised any more but ……. they are using a 1995 image for UHI adjustments to make ‘projections’ and ‘scenarios’ that will affect every country in the world.
How much longer do we have to wait for the politicians who are espousing world-saving schemes such as ETS, Cap-and-Trade, CPRS and other ‘solutions’ to realise that the ‘science’ is not keeping up?
Why is there no independent, non-political, non-parochial, open-source, verifiable, world-wide climate science organisation (this is obviously NOT the UN/IPCC) that tracks, reports and uses up-to-date climate information that can be tested, verified, validated, repeated, etc?
With the amount of money being spent to promote clean energy schemes, surely this is the FIRST thing that any politician would want to see before sending their country back to the Dark Ages.
Re the lights – in my experience, there are not many lights placed in locations that do not also have roadways, parking lots, buildings, etc. It’s not the energy used for the lights that is the issue, it’s the heat retention of the surrounding non-rural items that affects UHI. The lights are just an indicator that these UHI-affecting items are likely to exist.

September 3, 2009 4:43 pm

Further proof of the high temperatures during the medieval warm period is shown in this recently discovered brochure for an 11th century beach vacation in Greenland:

September 3, 2009 5:20 pm

SOME YEARS ago – say, 2005? 2006? – the National Geographic had a cover story using a satellite night time view of lights over the continental US to claim (what else?) proof of man-made global warming’s menace.
Along the lines of this thread, I thought back then that it served as evidence for the opposite.
Fortunately, thanks to Watts – many volunteers, and of course Roger Pielke, Sr – the hard evidence to evaluate which side is served by such a claim is soon to become available. Thus, a proper, scientific analysis can be made. It hasn’t been up to now.

September 3, 2009 5:40 pm

Measuring GDP and Trade is little more than a guess and has little basis in reality. Measuring money flows is even harder to measure than heat flows because there are no laws of Thermodynamics that must be obeyed.

September 3, 2009 5:40 pm

Using light intensity to indicate economic growth may not work in all areas. One place in particular, Manitoulin Island, has had a push in the last decade or so for replacements and new installations of lights outside to be compatible with their dark sky philosophy. I am curious to know how many places have begun implementing such policies. — John M Reynolds

Tim Groves
September 3, 2009 5:51 pm

Also OT:
The Return of the hHockey Stick!!
The BBC are giving this one their best shot:
Arctic temperatures are now higher than at any time in the last 2,000 years, research reveals.
Changes to the Earth’s orbit drove centuries of cooling, but temperatures rose fast in the last 100 years as human greenhouse gas emissions rose.
Scientists took evidence from ice cores, tree rings and lake sediments.
Writing in the journal Science, they say this confirms that the Arctic is very sensitive both to changes in solar heating and to greenhouse warming.
The 23 sites sampled were good enough to provide a decade-by-decade picture of temperatures across the region.
The result is a “hockey stick”-like curve in which the last decade – 1998-2008 – stands out as the warmest in the entire series.

September 3, 2009 7:31 pm

Hmmmm so by corollorary Earth Hour* decreases GDP!

September 3, 2009 8:10 pm

Bulldust (19:31:40) : Duh. That’s the point.

Dave vs Hal
September 3, 2009 8:29 pm

On the surface North Korea looks like a good place for measuring temps and looking at stars. But unfortunately aerosols from the lands of the big brown cloud will interfere.

September 4, 2009 12:29 am

Actually, using night lights to guage urbanization is not good science. The contrast between the Koreas is indicative of this, as the North has a good amount of industrialization, they just can’t afford to operate street lights at night.
In US, Europe, and other developed nations, their night lights are dimming due to changeover to new street lights that are narrow spectrum or otherwise are dimmer or narrow beamed to reduce sky glare. A changeover in a city would make it seem like it had become less urbanized but this would be false, and without a lot of data to know which communities are using which lighting technologies, the overall night light data is well nigh useless.

September 4, 2009 1:27 am

William (10:10:06) : “North Korea looks like the best example of what the warmers would like the rest of the world’s economies to look like.”
I believe it is called The Losers’ Revenge, William. Sad people; but carrying a lot of (or is that “an obscene amount” of) power in these days of bitterness raging.

Mark Fawcett
September 4, 2009 3:00 am

Tim Groves (17:51:15) :
Also OT:
The Return of the hHockey Stick!!
The BBC are giving this one their best shot:

Indeed they are, for what it’s worth I’ve sent the following email to the reporter (, not holding my breath…:
“Dear Mr Black,
I know you would probably rather eat Harkal (try it, I have, can’t say I recommend it) than visit a site that dares to challenge certain results but why don’t you nip over to and see what they have to say about this latest study. The guys there don’t, and I mean really don’t, have an anti-warming stance; they are however, world-class statisticians who can’t bear to see mathematics misused.
Once again they have found wanton bad practice with regard to proxy analysis. I notice in the article that there is, shock of shocks, another hockey-stick style graph. There aught to be a lesson taught in schools that says “kids – if you see this shape, you’ve probably gone badly wrong somewhere”.
Also, the study (or your report) manages to link the claimed warming to the shrinking of the Arctic sea-ice, this is simply wrong. Anyone with a modicum of nouse can discover that the 2007 minima was caused by unusual wind patterns; it is that driver that has by far the greatest effect on sea ice extent. Don’t believe me, go check up on it. It must be a shame that after the 2007 low-point, the Arctic sea ice has increased year on year and is now back to almost where it’s been on average over the last 10 years. It is a bigger shame that you don’t report that fact (as it’s a real observation, not a computer simulation or a mathematical manipulation) or the fact that Antarctic sea-ice is increasing year-on-year.
I hope, against hope, that one day a journalist like your goodself will remember that a cornerstone of good, honest, reporting is to take a balanced view and report on some of the obvious flaws in the arguments that are being repeatedly presented. The science is far from settled.
Anyway, best regards,
Mark Fawcett”

September 4, 2009 10:18 am

In so far as this works it suggests that electricity usage is a very good direct measure of GDP since almost all of this light will be electritc.
This has implications for the fact that China is now producing more electricity than the EU & growing fast. Shows how destructive western Luddism has been.

September 4, 2009 11:18 am

You might want to specifically start a program of looking at Airport growth over time. Don’t know if you could use the light data, but maybe (airports add every brighter lights as they get larger with faster airplanes that need to spot the airport further away).
In GIStemp, they use A LOT of airports and this will have a big effect. In fact, that single site that is used as a “rural” site to “correct” the most urban stations is an airport. I could not find it in database, though. But I have some pictures and data on it here:
If it can be mapped to a entry, I’d love to know. If it is a ‘missing station’, it probably ought to get some attention…

John S.
September 4, 2009 11:59 am

Thank you to all who responded to my quest for relatively uncorrupted temperature data in East Asia. I should have explained that it’s that region (along with Central Africa) that is most lacking in 100-yr long records from smaller towns, hence my focused interest in that region in conducting a world-wide survey. I had hoped that Anthony might point out a Chinese data source to update truncated GHCN records. Anybody out there in China who’s listening?

September 7, 2009 12:08 am

TonyB (15:29:32) :
Frank Lansner (13:36:13) : said
But… By now it actually seems that only 20% of the climate-visitors appears to be coming afterall…
(Leaving the authorities with a gigantic loss of money for the reservations..!)”
Can you make a block booking of 250 rooms at a very good price and get 250 delegate tickets?
Now, who fancies a trip to Copenhagen as part of the WUWT delegation?

Great isea. I wonder if Joe Bast could be talked into helping out with this.

September 8, 2009 5:06 am

John S.
That data is likely a state secret. People can lose their kidneys over releasing that sort of data.

%d bloggers like this: