What WorldView-3 is really great about
After reading the article about the WorldView-3 satellite published today (at the time of publication, the article was already removed in drafts), I first wanted to write a detailed comment to it, but then I realized that it would be longer than the article itself. Therefore, I decided to arrange it as a separate post with detailed explanations of what is actually new in the design of the sensors of the new WorldView-3 satellite except for a slightly better resolution than its predecessors. And the satellite really has several new devices at once in its design, which, after its launch, will allow solving remote sensing problems at a higher level.
Some simple facts
The WorldView-3 satellite belongs to DigitalGlobe , which has been supplying satellite imagery to consumers for many years, including well-known cartographic online services such as Google Maps, Yandex Maps, Bing. “Orbital telescope” (as it was called in the previous article) to call this satellite is not quite true, since astronomical satellites, which look not at Earth, but at other space objects, are called this term. WorldView-3 is thus the “Earth remote sensing spacecraft”.
About resolution
')
Indeed, the resolution of WorldView-3 sensors is slightly higher than the resolution of its predecessors - WorldView-1 and 2, devices of the QuickBird series. But it should be clearly understood that the different sensors installed on the satellite, this resolution is different.
The panchromatic sensor, which, in simple words, takes a black and white picture in the visible range of 450-800 nanometers (800nm ​​slightly goes into the infrared range) has the highest resolution. The company itself DigitalGlobe declares for it 31 centimeters per pixel for the vertical direction of shooting and 34 centimeters - with a deviation from the vertical of 20º. For comparison, WorldView-2 had a 46 centimeters panchromatic sensor resolution.
Color image satellites are obtained using the so-called multispectral sensors. These are sensors that are shot in separate narrow ranges, from which you can then assemble a so-called “composite”, including, among other things, a picture in natural colors. However, the resolution of multispectral sensors is usually lower than that of the panchromatic one. So here, the eight ranges of the multispectral sensor have a resolution with vertical shooting of 124 centimeters per pixel and 138 - with a deviation of 20º. The predecessor, these figures were 180 and 240 centimeters. Of course, when processing survey data, methods for improving resolution using panchromatic sensor data can be used; however, this only improves the ability to distinguish parts, but not color.
The resolution when deviating from the vertical is indicated separately intentionally: in practice, a rather large part of the images can be obtained at an angle, which reduces the final resolution by 10-15%.
In the case of WorldView-3, the set of sensors is not limited to these two types, which is a fundamentally new technical solution for high-resolution devices. On the characteristics of the third and fourth type, read on.
Spectral capabilities
As mentioned above, WorldView-3 has, like many other satellites, panchromatic and multispectral sensors.
Panchromatic covers with its range almost the entire visible eye spectrum from blue-violet to red, and also captures a little bit of the near infrared range.
The multispectral WV-3 sensor differs little from its capabilities from the WV-2 sensor, the difference is only in a slightly higher resolution, as was mentioned above. It has three channels, approximately corresponding to red, green and blue in RGB, an additional yellow channel, an extreme red channel, two near-infrared channels and a blue-violet channel for studying coastal waters. Total - 8 channels. The continuity of spectral characteristics is important for maintaining the methods that were developed using WorldView-2 data. Data width of all eight ranges - 11 bits.
I will briefly explain why this set is needed.
A fundamentally new sensor on WorldView-3 is the shortwave infrared radiation sensor (SWIR). Previously, such data was obtained by lower resolution satellites, such as Landsat. There, the resolution was only 30 meters per pixel, whereas in the WV-3 it is 3.7-4.1 meters with vertical shooting and a deviation of 20Âş. The SWIR sensor also has eight channels that cover the spectrum from 1195 to 2365 nanometers, but the range is not continuous, but covered by narrow bands. The choice of ranges is also associated with the properties of vegetation, atmospheric processes, and the four farthest from the visible ranges can be used in problems of geological research. The SWIR sensor data width is 14 bits.
Another innovative solution in the WorldView-3 sensor set is the CAVIS sensor, abbreviated as Cloud, Aerosol, Water Vapor, Ice, Snow (cloudiness, aerosols, water vapor, ice, snow). This sensor is relatively low, only 30 meters per pixel, of resolution, covering individual sample channels from 405 nm to 2245 nm. Its main purpose is to determine the state of the atmosphere and identify special cases of the earth’s surface that affect image quality. The set of channels is chosen so that, for example, white clouds can be distinguished from snow and ice, and various types of dust and smoke in the air can be distinguished from water vapor. A similar solution has already been used on the Landsat 8 satellite for cloud detection, but WorldView has a more advanced and versatile system that will allow for more accurate and effective correction of images.
Thus, the WorldView-3 satellite should indeed become, after its launch, a fundamentally new remote sensing tool, but not at all because of something like a fantastically greater resolution in the visible range.
As for the expectations that this satellite will help improve the coverage of map services, then it will certainly contribute. You do not need only to forget that map services save their money, and very rarely order the shooting of any area on purpose. They prefer to buy much cheaper archival images, that is, those that have already been ordered by any commercial or government consumers. Because some areas that are not interesting to these consumers, and remain without high resolution shooting for many years. So the quality of the Earth’s coverage of images on Google Maps or on Yandex does not directly depend on the number of remote sensing satellites.
Some simple facts
The WorldView-3 satellite belongs to DigitalGlobe , which has been supplying satellite imagery to consumers for many years, including well-known cartographic online services such as Google Maps, Yandex Maps, Bing. “Orbital telescope” (as it was called in the previous article) to call this satellite is not quite true, since astronomical satellites, which look not at Earth, but at other space objects, are called this term. WorldView-3 is thus the “Earth remote sensing spacecraft”.
About resolution
')
Indeed, the resolution of WorldView-3 sensors is slightly higher than the resolution of its predecessors - WorldView-1 and 2, devices of the QuickBird series. But it should be clearly understood that the different sensors installed on the satellite, this resolution is different.
The panchromatic sensor, which, in simple words, takes a black and white picture in the visible range of 450-800 nanometers (800nm ​​slightly goes into the infrared range) has the highest resolution. The company itself DigitalGlobe declares for it 31 centimeters per pixel for the vertical direction of shooting and 34 centimeters - with a deviation from the vertical of 20º. For comparison, WorldView-2 had a 46 centimeters panchromatic sensor resolution.
Color image satellites are obtained using the so-called multispectral sensors. These are sensors that are shot in separate narrow ranges, from which you can then assemble a so-called “composite”, including, among other things, a picture in natural colors. However, the resolution of multispectral sensors is usually lower than that of the panchromatic one. So here, the eight ranges of the multispectral sensor have a resolution with vertical shooting of 124 centimeters per pixel and 138 - with a deviation of 20º. The predecessor, these figures were 180 and 240 centimeters. Of course, when processing survey data, methods for improving resolution using panchromatic sensor data can be used; however, this only improves the ability to distinguish parts, but not color.
The resolution when deviating from the vertical is indicated separately intentionally: in practice, a rather large part of the images can be obtained at an angle, which reduces the final resolution by 10-15%.
In the case of WorldView-3, the set of sensors is not limited to these two types, which is a fundamentally new technical solution for high-resolution devices. On the characteristics of the third and fourth type, read on.
Spectral capabilities
As mentioned above, WorldView-3 has, like many other satellites, panchromatic and multispectral sensors.
Panchromatic covers with its range almost the entire visible eye spectrum from blue-violet to red, and also captures a little bit of the near infrared range.
The multispectral WV-3 sensor differs little from its capabilities from the WV-2 sensor, the difference is only in a slightly higher resolution, as was mentioned above. It has three channels, approximately corresponding to red, green and blue in RGB, an additional yellow channel, an extreme red channel, two near-infrared channels and a blue-violet channel for studying coastal waters. Total - 8 channels. The continuity of spectral characteristics is important for maintaining the methods that were developed using WorldView-2 data. Data width of all eight ranges - 11 bits.
I will briefly explain why this set is needed.
- With red, green and blue - everything is clear: they are needed to obtain images in natural colors, and green is also needed to measure the properties of vegetation.
- An additional yellow channel also makes it possible to assess the state of terrestrial and aquatic vegetation (algae), but to do this more accurately than if the data on reflected yellow would be obtained from a combination of green and red channels.
- The blue-violet channel serves for the study of coastal waters, since the light of this range penetrates deeply enough through the ocean water. Also, this range is absorbed by chlorophyll of healthy plants. Together with the yellow and red channels, this range allows you to identify flooded areas during natural disasters.
- Border red is designed to examine the health status of plants.
- The first near infrared range allows you to estimate the amount of water in both biomass and soil. Also used to highlight the image of reservoirs.
- The second near infrared has the same function, but is less susceptible to atmospheric distortion due to haze.
A fundamentally new sensor on WorldView-3 is the shortwave infrared radiation sensor (SWIR). Previously, such data was obtained by lower resolution satellites, such as Landsat. There, the resolution was only 30 meters per pixel, whereas in the WV-3 it is 3.7-4.1 meters with vertical shooting and a deviation of 20Âş. The SWIR sensor also has eight channels that cover the spectrum from 1195 to 2365 nanometers, but the range is not continuous, but covered by narrow bands. The choice of ranges is also associated with the properties of vegetation, atmospheric processes, and the four farthest from the visible ranges can be used in problems of geological research. The SWIR sensor data width is 14 bits.
Another innovative solution in the WorldView-3 sensor set is the CAVIS sensor, abbreviated as Cloud, Aerosol, Water Vapor, Ice, Snow (cloudiness, aerosols, water vapor, ice, snow). This sensor is relatively low, only 30 meters per pixel, of resolution, covering individual sample channels from 405 nm to 2245 nm. Its main purpose is to determine the state of the atmosphere and identify special cases of the earth’s surface that affect image quality. The set of channels is chosen so that, for example, white clouds can be distinguished from snow and ice, and various types of dust and smoke in the air can be distinguished from water vapor. A similar solution has already been used on the Landsat 8 satellite for cloud detection, but WorldView has a more advanced and versatile system that will allow for more accurate and effective correction of images.
Thus, the WorldView-3 satellite should indeed become, after its launch, a fundamentally new remote sensing tool, but not at all because of something like a fantastically greater resolution in the visible range.
As for the expectations that this satellite will help improve the coverage of map services, then it will certainly contribute. You do not need only to forget that map services save their money, and very rarely order the shooting of any area on purpose. They prefer to buy much cheaper archival images, that is, those that have already been ordered by any commercial or government consumers. Because some areas that are not interesting to these consumers, and remain without high resolution shooting for many years. So the quality of the Earth’s coverage of images on Google Maps or on Yandex does not directly depend on the number of remote sensing satellites.
Source: https://habr.com/ru/post/212673/
All Articles