Skip to Content

Google Earth Blog

Syndiquer le contenu
The amazing things about Google Earth
Mis à jour : il y a 21 min 3 sec

Take LaBeouf, Rönkkö & Turner anywhere

il y a 5 heures 45 min

Last week we talked about how American actor Shia LaBeouf had been tweeting a cryptic message using geographic coordinates that we expected would spell out “Take me anywhere”. It turns out that that guess was correct and that he meant it literally. Together with artists Nastja Säde Rönkkö & Luke Turner he will be tweeting a location where anyone can go and pick him up and take them anywhere. Once at a new destination, they will again tweet the coordinates and await a new lift. Follow their travels on this website.

Read more about it here

Interestingly, the same three artists have previously collaborated in another art project involving taking a different kind of lift

Here is the full message Shia tweeted:

To see the above letters in Google Earth download this KML file.

A little known feature of Google Earth:

Select the folder of placemarks and then click the “Play Tour” button to see all the placemarks in a tour.

The post Take LaBeouf, Rönkkö & Turner anywhere appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Animating Lusaka, Zambia

mer 25-05-2016

We were recently having a look at the historical imagery of Lusaka, Zambia, and we noticed that there is a lot of historical imagery for the last two years. There is an image nearly every week. Lots of historical imagery provides the ideal opportunity for animations, so here are two that we tried out.

The first animation uses about one image per month for two years. The idea was to try and show the changing of the seasons.

Speed in milliseconds per image:

The second shows a shopping mall (East Park Mall) being built. It consists of 134 images captured almost weekly.

Speed in milliseconds per image:

You can adjust the speed of the animations by dragging the slider below each animation.

It would be nice if Google Earth could do this sort of animation natively. It is possible to create a tour with historical imagery and we have in the past created tools to help do this. However, Google Earth cannot animate faster than about one image per second and that is really too slow for the above animations to look good. For comparison you can download the two tours here. Unless you have fast internet it could take a while for all the images to load and require a lot of download data.


The post Animating Lusaka, Zambia appeared first on Google Earth Blog.

Catégories: Sites Anglophones

3D imagery slow down?

mar 24-05-2016

As our regular readers know we maintain a map of areas that have Google’s 3D imagery. We not only track current coverage but also have a timeline. Google has been adding to the area covered with 3D at a fairly consistent rate for the last two to three years. Increasingly, however, the 3D imagery additions are updates to previously existing areas rather than new coverage. These are much harder to track, so we do not know just how much has been updated in this manner. Over the past month or so almost every update has been an update to a previously existing area with only minor extensions and occasionally deletions (which we largely do not bother to track).

3D area covered in square kilometres.

So is this the new normal? Google has covered most of the major population centres in the regions where they are most active (the USA and Europe). Other parts of the world still have a few major cities outstanding (Cape Town, South Africa for example). Collecting and processing 3D imagery must be quite an expensive exercise, so it is unlikely Google will aim to cover all the areas in between cities even for the USA. It is a pity they haven’t yet got a way of showing the dates the imagery was collected and there is no ‘historical 3D imagery’ option, so frequent updates would be rather a waste unless they are significantly improving the quality with each refresh.

The post 3D imagery slow down? appeared first on Google Earth Blog.

Catégories: Sites Anglophones

The Lost Mayan City that probably wasn’t

lun 23-05-2016

A couple of weeks ago there was this story about a fifteen year old Canadian boy who might have found a lost Mayan city. It all started as a school science project. Apparently he believed the Mayans may have built their cities according to star charts and using the locations of known Mayan cities and a map of constellations he found what he thought might be the location of a missing city. He then looked at satellite imagery of the location and saw some ‘linear features’, suggesting something might be there. It didn’t take long, however, before experts expressed the opinion that the Mayans did not build their cities according to constellations and that what could be seen in the satellite imagery was almost certainly just an old field.

The original story was a bit short on details, but we were able to identify the location shown in this article. There are only two images for the location in Google Earth. The default image is from 2006, and there is a lower resolution image from 1970.

In the 2006 image we can see that it is next to a lake and looks like the trees are shorter in a rectangular area. There are similar areas nearby and around the edges of the lake.

In the 1970 image it looks like the area was cultivated (1) and there are other cultivated areas nearby (2) and a very similar one near a lake (3).

To find the location above in Google Earth download this KML file.

Our conclusion from the imagery is the feature is a formerly cultivated field.

We think it reasonable to assume at this point that the boy did not successfully find a lost Mayan city, but he also didn’t deserve some of the criticism he received in the press. Quite a number of archaeologists and even amateurs have made significant finds with the use of Google Earth and other satellite imagery and we encourage the idea. It would seem the science fair judges agree and awarded him the gold prize for his efforts. The photo in this article (French) suggests there was quite a lot more to the project, which we do not have access to.

The post The Lost Mayan City that probably wasn’t appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Getting HIRISE imagery into Google Mars

ven 20-05-2016

We came across this interesting article which suggests that Mars may have not only had large seas, but that it may have experienced gigantic Tsunamis, the results of which can be seen to this day. If you want to see what oceans would have looked like on Mars, see this post.

The article mentions the use of THEMIS imagery. The THEMIS instrument is an infrared camera on the Mars Odyssey space craft. Google Mars has a number of Global maps, including ‘Daytime Infrared’ and ‘Nighttime infrared’, which we believe show THEMIS imagery. However, THEMIS has been capturing imagery for 8 years and we thought it might be interesting to see what imagery is available and whether it can easily be put into Google Earth.

A map of all the THEMIS imagery captured can be found here. It is very easy to use. On the left hand side select various parameters such as which year the imagery was captured, whether you want daytime or night time imagery and whether you are interested in visible imagery or infrared imagery (or both). Next, click on the location of interest and on each click, the map zooms in. When you are sure you have located the image you want, click the ‘Select’ button (arrow icon) from the tool bar above the map and click on the map again. It will then display the available images for that location below the map.

Finding THEMIS imagery.

Click one of the images shown and it opens in a new tab in the browser with details about the image and the option to download it in various formats.

We chose to use JPG. Next, create an image overlay in Google Mars using the downloaded image. We had some difficulty aligning the images, especially for long strips. The geographical features can look quite different in different wavelengths. Also, we couldn’t easily use the Google Mars ‘global maps’ as they appear above image overlays. When we finally managed to identify and align the ends of the strip, the middle was out of alignment. This suggests that for best results the image needs to be re-projected, which would require specialised knowledge and software.

To see a couple of images we tested with and location ‘C’ from the article about tsunami’s, download this KML file.

The post Getting HIRISE imagery into Google Mars appeared first on Google Earth Blog.

Catégories: Sites Anglophones

False colour imagery

jeu 19-05-2016

Google has recently done a minor imagery update. The only May image is a DigitalGlobe image of the Fort McMurray wildfire, although it does not include Fort McMurray itself. When we looked at the Fort McMurray wildfire a few days ago we noted that there is a wide variety of satellite imagery available. However, most of it is low resolution imagery not suitable for Google Earth and the high resolution imagery from DigitalGlobe is false colour imagery. We thought this might be a good time to make some observations about false colour imagery.

False colour imagery involves mapping different spectral bands to the colours visible to the human eye: red, green and blue. Read more about false colour imagery on Wikipedia and NASA’s Earth Observatory.

Earth observation satellites, such as Landsat and Sentinel, capture all their images in a number of bands and any image can be rendered in natural colour or various combinations of false colour. However, it would appear this is not the case for at least some of DigitalGlobe’s satellites as they often only have false colour imagery of certain events. False colour imagery is good for highlighting changes to vegetation caused by fire so DigitalGlobe had good reasons for capturing false colour imagery. But the fact that they don’t seem to have true colour imagery suggests that their satellites do not, by default, capture all wavelengths that they are capable of in every image. This is most likely due to data storage and transfer limitations on the satellites.

[ Update : We received feedback from DigitalGlobe (see the comments below) stating that they do, in fact, capture and store the visible wavelengths and if necessary could create true colour images for any location where they have false colour imagery. The reason why false colour imagery of Fort McMurray was shared is because infrared wavelengths penetrate smoke and highlight vegetation changes better than visible wavelengths (the false colour imagery we saw is most likely the Near Infrared, Red and Green bands.]

Google Earth has a surprising amount of false colour imagery. It is very common across the Sahara, Saudi Arabia and Indonesia. There are also some instances of it being used for particular events such as earthquakes and fires.

False colour greatly improves contrast in desert regions such as these sand dunes in the Sahara.

Sometimes it results in striking blues such as the above location in the Sahara that looks like the middle of the ocean.

False colour not only improves contrast but can distinguish between different types of sand, which in natural colour (the yellow areas) look nearly identical.

To find the above locations in Google Earth download this KML file.

The post False colour imagery appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Shia LaBeouf tweets cryptic message with GPS coordinates

mer 18-05-2016

American actor Shia LeBeouf has been tweeting a GPS coordinate daily since May 9th without any explanation. The coordinates are all near Denver Colorado. As pointed out in this article, the geographic features at each location resemble a letter of the alphabet. All the letters are clearly readable with north at the top of the screen, which has the benefit of being readable in Google Maps.

All images from Google Earth.

The message so far appears to read TAKE ME ANY

It is believed the full message may be TAKE ME ANYWHERE. LaBeouf is expected to appear at the Boulder Museum of Contemporary Art as part of a panel discussion with artists Nastja Säde Rönkkö and Luke Turner on May 22 as part of the MediaLive festival and the message may be a reference to artist Nastja Säde Rönkkö’s project “Take me anywhere”.

This is not the first time someone has used geographic features to spell things out. We have in the past looked at a font created from satellite images and there is that lets you send messages with letters in satellite images.

The post Shia LaBeouf tweets cryptic message with GPS coordinates appeared first on Google Earth Blog.

Catégories: Sites Anglophones

An aerial imagery oddity

mar 17-05-2016

We recently had a query from a GEB reader about the imagery near Harrisburg, Pennsylvania, and we noticed an oddity that we thought worth sharing. The same aerial imagery has been obtained from three different sources, processed in three different ways and assigned three different dates. This highlights the fact that the dates on aerial imagery from third parties should not be relied upon.

The first image is dated April 6th, 2003 and attributed to “PA Department of Conservation and Natural Resources-PAMAP/USGS”.

The second image has no attribution and is just dated 2003. When imagery in Google Earth is assigned a date range (or, as in this case, just a year), then the start date of the range is shown in the status bar at the bottom and the end date is shown on the timeline at the top left.

The third image is dated February 27, 2004 and attributed to “U.S. Geological Survey”.

We chose a busy parking lot to show that the cars are identical in all three images, indicating that they all come from the same original image. The first and third images have similar colouring, whereas the second has brighter colours.

Zooming in, we see that the third image looks a bit more pixelated and has a tiny bit more detail. We believe the first and second have been processed to reduce the pixelation, but at the expense of detail.

We believe the original image was captured by the U.S. Geological Survey (USGS) then shared with the PA Department of Conservation and Natural Resources and possibly another third party. Then Google obtained the imagery from those third parties and directly from the USGS and didn’t realise it was in fact the same image. It is likely the imagery was obtained as part of large batches of imagery.

We believe the dates on more recent aerial imagery gathered by Google is more accurate. Certainly, we were able to confirm the dates on the imagery of the recent earthquake in Japan that we looked at last week was correctly dated, as we knew the exact date of events visible in the imagery.

The post An aerial imagery oddity appeared first on Google Earth Blog.

Catégories: Sites Anglophones

The Fort McMurray Wildfire

lun 16-05-2016

On May 1, 2016, a wildfire started near Fort McMurray, Alberta, Canada. On May 3, According to Wikipedia, it swept through the community, destroying more than 2,900 homes and buildings and forcing the largest wildfire evacuation in Alberta’s history.
.sliders img{max-width:none; }

Google Earth does not yet have any imagery of the event, but various imagery providers have captured and shared imagery.

Terra Bella
Google’s Terra Bella (formerly Skybox Imaging) captured the event and the imagery has been shared via this map on Google Crisis Response.

Google Crisis Response page for the Fort McMurray wildfire.

Be sure to check the layers for different dates.

You can download Terra Bella imagery here. It is shared with the CC-BY-SA which allows you to do what you like with it on condition you give appropriate attribution.

May 3rd and May 5th, Beacon Hill, Fort McMurray. Terra Bella imagery.

To see the above images in Google Earth download this KML file.

Planet Labs
Planet Labs has captured imagery of the event. Read more about it on their blog. There are a number of ways to download the imagery. We chose to download the three zip files from this page totalling 4 Gb. Each zip file contains a number of tiles, which can be drag and dropped into Google Earth Pro with ease.

Planet Labs imagery showing Fort McMurray on May 6th, 2016.

For a low resolution version of the May 6th imagery from Planet Labs download this KML file (17 Mb).

Airbus Defence & Space
Airbus Defence & Space has an image in its gallery here, which it says was captured by the SPOT 6 or SPOT 7 satellite. You can download it from that page to put into a Google Earth overlay. However, it does not cover Fort McMurray itself and is about 20 km to the east.

Global News has this article which features more Airbus Defence & Space imagery, this time captured by the Pleiades satellites.

DigitalGlobe has captured some false colour imagery, which is featured, with before and after comparisons, in this article and on the DigitalGlobe blog

The Wikipedia page on the wildfire features a Landsat 7 image showing the characteristic stripes we talked about last week.

Landsat 8 has an image captured on May 15th that clearly shows the burn scar.

Landsat 8 image showing the burn scar.

To see it in Google Earth download this KML file (low resolution and cropped).

jQuery( document ).ready(function() { jQuery(function(){jQuery('#slider1').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); jQuery(function(){jQuery('#slider2').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); jQuery(function(){jQuery('#slider3').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); });

The post The Fort McMurray Wildfire appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Seeing worm poo from space

ven 13-05-2016

We recently came across this interesting article about a South American phenomenon of mounds known as surales. Apparently scientists have demonstrated that they consist largely of earthworm excrement and are created by the earthworms. The patterns they produce are reminiscent of fairy circles and ant and termite cities that we have covered in the past.

We tried looking for the surales, but they proved quite difficult to find. We found places we thought might be surales but had no way of confirming they were not termite mounds or just small bushes. In one case we even found a hopeful location with Street View but we could see termite mounds in the Street View.

Possibly a mix of surales and termite mounds

Luckily, we came across this article, which includes a video of one of the scientists using Google Earth to show off some of the locations.

Surales come in a variety of sizes:

It is difficult to distinguish them from termite mounds, and sometimes they occur together:

Surales on the left, termite mounds on the right.

Surales form a variety of patterns:

The original science paper can be found here

For the above locations and more from the video, download this KML file.

The post Seeing worm poo from space appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Google Earth Imagery Update – Cyclone Fantala

jeu 12-05-2016

As we mentioned in yesterday’s post, Google has recently pushed out yet another imagery update.
.sliders img{max-width:none; }

April imagery – Green: as of April 17th, Blue: as of April 21st, Red: As of May 10th

March imagery – Green: as of April 17th, Blue: as of April 21st, Red: As of May 10th

To see the outlines in Google Earth download this KML file. Note that the actual imagery is smaller than the outlines.

We have found quite a lot of DigitalGlobe imagery (with the help of their First Look map) that was captured because of specific events. However, in many cases we were not able to see anything of interest in the imagery. The locations are noted in the KML file at the end of this post. We did find some places of interest:

In mid April Cyclone Fantala struck the Farquhar Group of islands, which are part of the Seychelles.

Almost all the buildings on the island are completely destroyed. Apparently the residents had been evacuated.

In Kerala, India, a large explosion due to a fireworks accident at Puttingal Temple killed 114 people and injured 350 others.

Puttingal Temple, Kerala, India, before and after explosion.

The Five-hundred-meter Aperture Spherical Telescope (FAST) in Guizhou Province, China is nearing completion. We previously featured it when looking at radio telescopes

Five-hundred-meter Aperture Spherical Telescope (FAST), Guizhou Province, China.

To see the locations featured above as well as a number of locations where imagery has been captured for a specific reason, download this KML file.

jQuery( document ).ready(function() { jQuery(function(){jQuery('#slider1').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); jQuery(function(){jQuery('#slider2').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); });

The post Google Earth Imagery Update – Cyclone Fantala appeared first on Google Earth Blog.

Catégories: Sites Anglophones

The Kumamoto earthquakes

mer 11-05-2016

Google has been pushing out imagery updates almost weekly recently. The latest update includes imagery from Kumamoto, Japan, which suffered a series of large earthquakes in April, 2016. Google has managed to capture aerial imagery not just after the event, but during the event. For example, we can see aerial imagery of the slopes of Mt Aso captured on April 15th, and again on April 16th showing that a large number of landslides took place between the images being captured.
.sliders img{max-width:none; }

Before and after of multiple landslides on Mt Aso, Japan.

Before and after of a landslides that swept over some houses.

A bridge and road destroyed by a landslide.

Some of the landslides channelled together to produce a mudslide at the bottom:

Mudslide caused by landslides on Mt Aso, Japan.

One hillside didn’t quite collapse, but looks very unsafe.

Large cracks across a hillside that has only slid a little bit.

The earthquakes started with a foreshock on April 14th, which struck the eastern Kumamoto suburb of Mashiki. Google was very quick to respond, as the aerial imagery is captured the following day, April 15th. The largest shock took place on April 15th, at 16:25 UTC, which is April 16th, 01:25 JST, so, after the first Google aerial image and before the second one.

Looking at Mashiki, we can see what looks like tiles missing in the roofs of buildings. It is difficult to tell just how much damage has been done. A look at this article, however, shows that many houses collapsed vertically with their roofs largely intact and the real damage was severe.

Collapsed houses in Mashiki, Japan.

In the city of Kumamoto, a bullet train was derailed, but apparently it was not carrying passengers and nobody was hurt. The imagery is dated 15th April, so it appears it was derailed by the foreshock.

Derailed bullet train.

Kumamoto Castle appears to have suffered some damage from the foreshock and then more serious damage in the main quake. There are some relatively poor quality satellite images from after the main quake.

To see the above locations and many more in Google Earth download this KML file.

jQuery( document ).ready(function() { jQuery(function(){jQuery('#slider1').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); jQuery(function(){jQuery('#slider2').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); jQuery(function(){jQuery('#slider3').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); });

The post The Kumamoto earthquakes appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Street View comes to Kyrgyzstan

mar 10-05-2016

Thank you to GEB reader ‘ecksemmess’ for letting us know that Kyrgyzstan has just received Street View.

Kyrgyzstan Street View coverage.

The last major Street View addition was Sri Lanka in late March so the map below shows the additions since then.

Street View additions in red, previously existing Street View in blue. Large version.

The lake Issyk Kul is the tenth largest lake in the world by volume. See in Street View.

Do you prefer your tandem bike inline or parallel? See in Street View.

If you look carefully you will notice both the snow and the trees cling to one face of each ridge while the other face is bare. See in Street View.

The post Street View comes to Kyrgyzstan appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Why there are stripes in the snow on Greenland

lun 09-05-2016

Recently, GEB reader Dan Jacobson, who lives in Taiwan, pointed out some large stripes in the snow on Greenland as seen below:

Stripes in the snow near the northern coast of Greenland.

Zoom out a bit and you can see the whole coastline is stripey and has square steps in it.

Stepped coastline.

We immediately recognised what this was, because we have seen it before. It is Landsat 7 imagery. If you download a Landsat 7 image captured after May 31, 2003, it looks like the image below. There is a clear strip down the centre, but the rest of the image has long spikes going out to the sides.

A Landsat 7 image.

The Landsat 7 satellite does not take a single photo of the whole scene that you see when you download a Landsat image. Instead it captures a small strip of imagery which is moved backwards and forwards across the direction of movement of the satellite in a zigzag pattern. This allows a single, low resolution camera to capture a large, high resolution image. We are not certain, but it would appear that Landsat 7 has a single sensor which captures the image one pixel at a time in a variety of wavelengths. A series of rotating mirrors are used to capture light from different directions to build up the image. Two of the mirrors, called the Scan Line Corrector (SLC), compensate for the forward motion of the satellite, thus changing what would otherwise be a zigzag into parallel lines. Landsat 7 was launched on April 15, 1999, and on May 31, 2003, the SLC failed, so what we see in Landsat 7 imagery captured after that date is the zigzag without that correction. Read more about it on the USGS website and Wikipedia.

To see a sample Landsat 7 image of Greenland in Google Earth, download this KML file. Keep in mind that it is a very low resolution version to keep the file size to a minimum.

Google could update Google Earth with better quality Landsat 8 imagery, or even Sentinel imagery, but since not many people look at the northern coast of Greenland, they don’t seem to have made it a high priority.

The post Why there are stripes in the snow on Greenland appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Controlling maps in the days of satellite imaging

ven 06-05-2016

We recently came across this story. Apparently India is considering enacting laws which would require companies providing mapping data to get a licence and approval before disseminating maps of any kind within India. The government appears to have two major concerns. One is the correct depiction of India’s borders (as claimed by India in the case of disputed borders), and the other is a security concern about showing sensitive sites on maps.

Almost every country has had some form of censorship laws with regards to maps, but most have largely given up due to the proliferation of maps and imagery that they have no control over.

We have in the past looked at censored imagery in Google Earth, and in most cases it is aerial imagery, which a country does have legal control over. However, most obviously censored locations have alternative uncensored imagery and the censorship only serves to draw attention to the location. Censoring satellite imagery is difficult, but not impossible. If you look at the maps of imagery updates over the last year then you will notice that Iraq, Afghanistan, Ukraine and Syria have not had any updates in the past year. We have also noticed that the Spratly Islands in the South China Sea also have not received any updates since they started to get attention due to China building artificial islands there. We do not know if this is a case of censorship or not. Hawaii, has also not received updates in the same time period and we can think of no reason why that would be censored.

Several countries have apparently censored 3D imagery, most noticeably, Greece, where Airports do not receive 3D coverage. But they do have satellite imagery and in some cases Street View.

The 3D imagery around Larrisa, Greece, clearly avoids the airport.

China has strong laws about mapping and insists on its maps being offset slightly from GPS coordinates for security reasons. In this they have been largely successful, but they have no control over satellite imagery when viewed outside China, so that is correctly aligned.

Disputed borders are often displayed differently, depending on which country you view them from in order to comply with the laws of the countries in question and the same may apply to disputed names of geographic features.

The end result of all this is that censorship laws on maps typically only affect the citizens of the country in question as the laws are unenforceable outside the country’s borders.

The post Controlling maps in the days of satellite imaging appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Sentinel vs Landsat imagery

jeu 05-05-2016

Yesterday we had a look at Snapsat, a useful website for obtaining Landsat imagery. The location we chose was Dallas, Texas and a track made by a tornado in December last year.
.sliders img{max-width:none; }

We thought it would be interesting to compare the Landsat imagery with the European equivalent – Sentinel imagery.

Landsat 8 captures colour images at 30 m per pixel, but also captures has a panchromatic band at 15 m per pixel, which can be combined (using a process known as pansharpening) with the colour bands to essentially achieve close to 15 m per pixel resolution. Sentinel 2A, on the other hand, captures images in colour at 10 m per pixel and does not use pansharpening. So the Sentinel imagery should be slightly better quality and we found this was indeed the case.

We are still learning the best ways to process imagery and neither image has been processed ideally.

Landsat 8 imagery vs Sentinel 2A imagery.
Images courtesy of USGS/NASA Landsat and Copernicus Sentinel data 2016.

To see the images in Google Earth download this KML file.

jQuery( document ).ready(function() { jQuery(function(){jQuery('#slider1').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); });

The post Sentinel vs Landsat imagery appeared first on Google Earth Blog.

Catégories: Sites Anglophones

Snapsat Beta for Landsat imagery

mer 04-05-2016

We have previously had a look at how to get Landsat imagery into Google Earth. We have used a number of different techniques over time, but in all cases, obtaining high quality Landsat imagery required a large download (typically over 700 Mb) and several stages of image processing to combine multiple bands into a full colour image.

We recently discovered Snapsat, a website dedicated to making it easier to obtain processed Landsat imagery. In March last year, Amazon announced the availability of Landsat data on Amazon Web Services (AWS). Snapsat builds on the AWS, offering to provide easy access and processing of the Landsat data. It appears to have been set up mid last year not long after the Landsat data was made available on AWS but not much work appears to have been done on it since then. However, it still works well and makes obtaining processed Landsat imagery remarkably easy. Learn how to use it here.

As we mentioned in this post there was a Tornado which struck Dallas, Texas in December last year. There is no imagery yet in Google Earth showing the main path of destruction. So we tried out Snapsat by downloading a Landsat image of Dallas from January 12th. The process was straight forward and only required a download of 107 Mb as opposed to the 772 Mb required to download the same imagery from Earth Explorer. And we can see the track of the tornado.

Tornado track, Dallas, Texas, as seen in Landsat imagery. Ends of the track marked with arrows.

There are, however, a few disadvantages to using Snapsat. The website does not do pansharpening, a process whereby a higher resolution grey-scale image is used in combination with the three colour bands to create a higher resolution image overall. This means that the imagery obtained via Snapsat is not the highest quality possible. Another issue is that the downloaded file does not include Geo-coding information so it cannot be automatically imported into Google Earth Pro. However, since we usually have to crop the imagery anyway, manual positioning is often necessary and not very difficult.

We got the same image from Earth Explorer and used GIMP to process the imagery (very inexpertly) and the result is noticeably sharper because of the pansharpening.

Tornado track, Dallas, Texas, as seen in Landsat imagery, with basic pansharpening via GIMP

To see the two overlays in Google Earth download this KML file.

The post Snapsat Beta for Landsat imagery appeared first on Google Earth Blog.

Catégories: Sites Anglophones

What’s that image: Earthquake and floods

mar 03-05-2016

Yesterday we mentioned that Google had pushed out an imagery update towards the end of April. Today we are having a look at some of the sights we have found so far.
.sliders img{max-width:none; }

On April 16th, 2016, Ecuador experienced an earthquake of magnitude 7.8. Google Earth has three satellite images captured the day after. All three images seem to have a bug where you cannot see the previous image in historical imagery but have to go back two dates to see prior imagery. Perdenales, Ecuador was not far from the epicentre of the earthquake and a number of buildings collapsed. The satellite imagery is not very high quality, but the devastation is visible if you look closely. To identify collapsed buildings, look for rubble in the streets rather than at the buildings, then compare with older imagery. See the fourth image in the slideshow in this story to see an aerial view.

Collapsed buildings in Perdenales, Ecuador.

Further south, in the town of Portoviejo we found a collapsed building as well as signs of minor flooding.

Collapsed building in Portoviejo, Ecuador.

The southern United States seems to have experienced a lot of flooding lately. Earlier this month we had a look at the flooding of the Mississippi River, which took place in January this year. There are now a number of images captured in mid March that show flooding in Texas and Louisiana.

The Sabine River floods Deweyville, Texas.

The Red River in flood near Coushatta, Louisiana.

The Ouachita River, Louisiana.

There are also satellite images of Austin and Houston, Texas, captured on March 12th. We are not sure if they relate to flooding and have not yet identified anything unusual in the images.

There has been further flooding in April, but there is not yet any relevant imagery.

For the locations featured in this post download this KML file. We have included in the KML some data from the USGS showing the epicentre of the quake and the area of impact.

jQuery( document ).ready(function() { jQuery(function(){jQuery('#slider1').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); jQuery(function(){jQuery('#slider2').beforeAfter({imagePath: '/js/utils/',showFullLinks : false});}); });

The post What’s that image: Earthquake and floods appeared first on Google Earth Blog.

Catégories: Sites Anglophones

A year of imagery updates in Google Earth

lun 02-05-2016

As we mentioned last week, Google did at least four imagery updates in the month of April. It takes long time to do imagery update maps and since imagery updates include imagery over a number of different months it could take up to a week to do maps for the last few months. So we thought of changing our algorithm to do them all at once. This is faster overall but still takes days to complete. We also decided to do the last twelve months, since it wasn’t too much extra trouble to do so. We are not yet done with the high resolution maps, but we can give a good idea of where updates were done over the past year.

Imagery updates over the past year – May 2015 to April 2016

The above animation uses NASA’s Blue Marble images taken from the animation that Frank did back in 2006. We did not include the poles in our imagery update maps, so there may have been some imagery updates for Antarctica and northern Greenland that we have missed.

Notice how the imagery avoids the snow line in the northern hemisphere. The pattern in the tropics is not as clear as the map we did last week showing how seasonal Google Earth imagery is.

You can also see the animation in Google Earth using this KML file.

For best results:

  • adjust the timeline control so that the two parts of the slider are snapped together.
  • in the timeline control settings dialog, set the animation speed to maximum, and tick ‘loop animation’.
  • press the ‘play animation button’ on the timeline control.
  • you may have to wait for a few loops of the animation for all the images to load properly.

The post A year of imagery updates in Google Earth appeared first on Google Earth Blog.

Catégories: Sites Anglophones

The best of Google Earth for April 2016

ven 29-04-2016

Google has pushed out new imagery at least four times this month. 1, 2, 3, and a very recent update that we haven’t covered yet!

We found quite a lot of interesting sights in Google Earth imagery this month, including:
* Nyssa, a very large painting by Jorge Rodriguez-Gerada.
* Super Bowl celebrations in Denver Colorado.
* Mississippi flooding.
* A collapsed building due to an earthquake in Taiwan.
* A train crash in Germany.
* Flooding in India, Tanzania and Brazil.
* Demolished buildings in Zaria, Nigeria, indicating a possible cover up of mass killings.

We did a series of posts attempting to estimate the size of the Google Earth database and how much space the different types of imagery require. First we looked at 3D imagery, then different types of 2D imagery, moved on to historical imagery and finally put it all together and came up with a very rough estimate of 3 Petabytes!

We created a map showing the most popular times of year for capturing imagery in different parts of the world. The resulting pattern was very interesting. In the extreme latitudes, there is a lot less imagery captured in the winter because of snow cover and, as GEB reader Chris pointed out, the shorter days and longer shadow make gathering good quality imagery more difficult. In the lower latitudes the situation is reversed, with more imagery gathered in the winter, most likely because there is typically more cloud cover in the summer months.

We had a look at some scratch marks on the floor of the Caspian Sea and compared the view from Landsat imagery and ASTER imagery. We forgot to mention that the area also has DigitalGlobe imagery from 2005, which also shows scratch marks.

We had a look at the Sentinel missions and hope to cover them more in later posts when we have learned how to process the imagery. Thank you to GEB reader ‘mzuehlkeMarco’ for pointing us to a tool called SNAP for processing the imagery.

We had a look at some examples of land reclamation that can be seen taking place in Google Earth imagery.

We had a look at some strange buildings in the desert and decided they are Surface to Air Missile (SAM) sites, which it turns out are remarkably common and easy to identify in imagery once you know what to look for.

At the end of March, Tesla unveiled the Tesla Model 3 and we had a look at Lithium mines around the world, which may supply the necessary Lithium for the Tesla cars.

The post The best of Google Earth for April 2016 appeared first on Google Earth Blog.

Catégories: Sites Anglophones