Water Vapor Imagery

Read...

Our look at visible and infrared imagery has hopefully shown you that using a variety of wavelengths in remote sensing is helpful because this approach gives us a more complete picture of the state of the atmosphere. Meteorologists can use visible and infrared imagery to look at the structure and movement of clouds because these types of images are created using wavelengths at which the atmosphere absorbs very little radiation (so radiation reflected or emitted from clouds passes through the clear air to the satellite without much absorption). Now, what if we took the opposite approach? What if we looked at a portion of the infrared spectrum where atmospheric gases (namely water vapor) absorbed nearly all of the terrestrial radiation? What might we learn about the atmosphere? Water vapor imagery addresses that question.

In case you didn't catch it in the paragraph above, let me be clear: Water vapor imagery is another form of infrared imagery, but instead of using wavelengths that pass through the atmosphere with little absorption (like traditional infrared imagery, which utilizes wavelengths between roughly 10 and 13 microns), water vapor imagery makes use of slightly shorter wavelengths between about 6 and 7 microns. As you can tell from our familiar atmospheric absorption chart, these wavelengths are mostly absorbed by the atmosphere, and by water vapor in particular. Therefore, water vapor strongly emits at these wavelengths as well (according to Kirchoff's Law). Thus, even though water vapor is an invisible gas at visible wavelengths (our eyes can't see it) and at longer infrared wavelengths, the fact that it emits so readily between roughly 6 and 7 microns means the radiometer aboard the satellite can "see" it.

This fact makes the interpretation of water vapor imagery different than traditional infrared imagery (which is mainly used to identify and track clouds). Unlike clouds, water vapor is everywhere. Therefore, you will very rarely see the surface of the earth in a water vapor image (except perhaps during a very dry, very cold Arctic outbreak). Secondly, water vapor doesn't often have a hard upper boundary (like cloud tops). Water vapor is most highly concentrated in the lower atmosphere (due to gravity and proximity to source regions like large bodies of water), but the concentration tapers off at higher altitudes.

The fact that water vapor readily absorbs radiation between roughly 6 and 7 microns also raises an interesting question -- just where does the radiation that ultimately reaches the satellite originate from? The answer to that question is the effective layer, which is the highest altitude where there's appreciable water vapor. In other words, the effective layer is the source for the radiation detected by the satellite. Above the effective layer, there is not enough water vapor to absorb the radiation emitted from below, nor is there enough emission of infrared radiation to be detected by the satellite. Any radiation emitted below the effective layer is simply absorbed by the water vapor above it.

In our previous discussion of traditional infrared imagery, I'm not sure if you realized that the radiation detected by the satellite only came from one distinct level in the atmosphere at a given point. If the column was clear, then the surface was detected; however, if the column contained clouds, then only the top-most layer of clouds was observed. The surfaces that emit the radiation that the satellite "sees" (highest cloud tops or the ground in the case of traditional IR imagery) are the "effective layers." A universal property of an effective layer is that only emissions from this layer are observed by the satellite. For a visual, consider emissions at a representative wavelength useful for traditional infrared imagery (10.7 microns, for example) from a cloudy atmospheric column (toward the left on the schematic below).

At traditional infrared wavelengths (like 10.7 microns), the satellite either sees radiation from the ground or the tops of clouds (left). The level from which the satellite observation is derived is called the effective layer. For water vapor imagery (right), the effective layer is defined as the highest level of appreciable water vapor whose radiation can be detected by the satellite. As with traditional IR imagery, all radiation emitted below the effective layer is absorbed and does not reach the satellite.

Credit: David Babb

In the column with clouds, radiation emitted from the top of the cloud reaches the satellite because no appreciable liquid water or ice exists above the cloud, giving the radiation a "free pass" to the satellite. Below the observed cloud layer (that is, the effective layer), any emissions from liquid water and ice are absorbed by the cloud layer that lies above them. Of course, if the air column is free of clouds, then the ground is the effective layer at longer infrared wavelengths, because the emissions that the satellite radiometer sees are coming from the ground (column farthest to the left in the graphic above).

Now let's carry this idea over to water vapor imagery (refer to the right portion of the above schematic). At the wavelengths used for water vapor imagery (between roughly 6 and 7 microns), water vapor very effectively absorbs and emits radiation. Another way to think about it is that at a wavelength like 6.7 microns (the sample wavelength used in the schematic), water vapor radiates just like liquid water and ice do at 10.7 microns. So, water vapor is an invisible gas at visible wavelengths and longer infrared wavelengths, but it "glows" at wavelengths around 6 to 7 microns.

The bottom line is that, for water vapor imagery, the effective layer lies in the uppermost region of appreciable water vapor. That is, the effective layer is the source region for the radiation detected by the satellite, as the columns toward the right in the schematic above show. Above the effective layer, there is not enough water vapor to generate a signal to be observed by the satellite. And as with clouds in the traditional IR example, any radiation emitted below the effective layer is simply absorbed by the water vapor above it. Therefore, the satellite measures the radiation coming only from the effective layer, and like traditional infrared imagery, this radiation intensity is converted to a temperature. Here's the key point: Water vapor imagery displays the temperature of the effective layer of water vapor (notice that the water vapor image below has a temperature scale, just like traditional infrared imagery). Commonly, water vapor imagery uses shades of gray, with warmer (lower) effective layers shown as dark and colder (higher) effective layers shown in white. As with traditional IR imagery, color enhancements are often added to identify key temperatures. In the image below, warm colors (oranges, reds) are added to emphasize warm effective layers, while cool colors (purples, blues, greens) are used to illustrate very cold effective layers.

Common water vapor imagery uses shades of gray with higher brightness temperatures rendered as dark and lower brightness temperatures shown in white. As with traditional IR imagery, color enhancements are added to identify key temperatures. In this image, warm colors are added to emphasize warm effective layers, while cool colors are used to illustrate very cold effective layers.

Credit: University Corporation for Atmospheric Research (UCAR)

You may hear on television or see other online explanations that suggest water vapor imagery measures the water vapor content of the atmosphere, but that's not really true. We can infer certain things about the moisture profile of the atmosphere based on the temperature of the effective layer, but the satellite isn't actually measuring the amount of water vapor present in order to create water vapor images.

So what can we infer by knowing the temperature of the effective layer? Just like traditional infrared imagery, we make the assumption that temperature decreases with increasing altitude, which implies that colder effective layers reside higher in the atmosphere. This means that if we know the height of the effective layer, we can infer the depth of the dry air above it (remember that we cannot make any assumptions about what lies below the effective layer). For example, consider the region shaded orange over southern California in the image above. Here the temperature of the effective layer is a relatively balmy -18 degrees Celsius. This temperature corresponded to a height of 19,000 feet (5.8 km) on this date -- approximately in the middle region of the troposphere (I looked up the temperature profile for this location and time). So, if the effective layer is located at 19,000 feet, then we can infer that some of the mid-level and all of the upper-level atmospheric column is dry. 

For a location over Denver, Colorado, which shows a temperature of - 30 degrees Celsius, the height of the effective layer was nearly 30,000 feet (9.1 km) on this date, and we can conclude that the upper troposphere contains more water vapor here than over southern California. Finally, turn your attention to the region of -60 degrees Celsius over central Texas (light blue). Here, the effective layer was way up at 40,000 feet (12.2 km) on this date -- at the boundary between the troposphere and stratosphere. Such a cold, high effective layer can only be caused by high ice clouds typical of the tops of cumulonimbus clouds. I should point out that at such low temperatures very little water exists in the vapor phase. However, ice crystals also have a fairly strong emission signature between 6 and 7 microns, so if you see such cold effective layers (less than -45 degrees Celsius or so), you are most likely looking at ice clouds (cirrus, cirrostratus, cumulonimbus tops, etc.) rather than at water vapor.

In the case above, did you notice that the lowest effective layer that we observed was 19,000 feet? That's not uncommon. Because emissions from water vapor near the earth's surface are absorbed by water vapor higher up, it's often impossible to detect features at very low altitudes. In other words, low clouds (stratus, stratocumulus, nimbostratus, and fair weather cumulus) are rarely observable on water vapor imagery. To see what I mean, check out the pair of satellite images below (infrared on the left, water vapor on the right). The yellow dot represents Corpus Christi, Texas, which was shrouded in low clouds (gray shading on the infrared image -- check out the meteogram for Corpus Christi). Now examine the water vapor image. The effective layer resided in middle troposphere as evidenced by the dark shading on the water vapor image (indicating a warm effective layer). However, not even a hint of the low clouds can be seen. In this case, the effective layer (located above the low clouds) absorbed all of the radiation emitted from below, rendering the low clouds undetectable on the water vapor image. For another example of low clouds not appearing on water vapor imagery, check out the Case Study section below.

An infrared image (left) at 10Z on January 14, 2003, shows a blob of low clouds (in gray) over the western Gulf of Mexico and the Texas Seaboard. The surface observation at Corpus Christi (yellow dot) is consistent with the IR satellite data. But there are seemingly no clouds visible in the water vapor image (right). The dark shading on the water vapor image indicates that the effective layer lies in the mid-troposphere (above the low clouds). Therefore, the radiation emitted by liquid water and water vapor in the tops of the low clouds was absorbed by water vapor higher up and never reached the satellite.

Credit: NOAA

If you look back carefully at our familiar atmospheric absorption spectrum, notice that absorption (and therefore emission) by water vapor isn't uniform in the range of wavelengths used for water vapor images (roughly 6 to 7 microns). Indeed, toward the higher end of the range, absorption is less than 100 percent, and using the different absorption and emission properties of water vapor near 7 microns allows satellites to "see" effective layers in different layers of the troposphere. Therefore, you'll sometimes find water vapor images called "upper-level", "mid-level" or "lower-level" water vapor images. While the altitude of the effective layer on any of these images varies based on the amount of water vapor in an air column (and how it's distributed), make sure that you're not fooled by these names. Even "lower-level" water vapor imagery typically detects effective layers between roughly 7,500 feet and 18,000 feet. In other words, most often, you're looking at emissions from effective layers of water vapor in the middle or upper troposphere on any of these satellite products.

Therefore, even "lower-level" water vapor imagery still can't often detect surface water vapor or the presence of low clouds. For example, check out this side-by-side comparison of a visible image and "lower-level water vapor" image from October 28, 2019. On this water vapor image, shades of yellow and orange mark regions with a warmer effective layer. Note that the lower-level water vapor image provides no indication of the presence of low clouds whatsoever (especially notable over Illinois and Indiana), because their tops were located below the effective layer at this time (their emissions were absorbed by water vapor higher up). The bottom line is that even on "lower-level" water vapor images, you cannot see near surface water vapor, fog, or low clouds, unless the atmospheric is extremely dry higher up (which is only possible in very cold, dry Arctic air).

Much like smoke from an extinguished candle, water vapor imagery helps forecasters trace mid- or upper-level winds.

Credit: Candle smoke / The Ewan / CC BY 2.0

Now that we've discussed how to interpret water vapor imagery, what might we use it for? Forecasters most often use water vapor imagery to visualize upper-level circulations in the absence of clouds. This is because water vapor is transported horizontally by high-altitude winds and thus can act like a tracer, much like smoke from an extinguished candle (as in the photo on the right). Consider this enhanced IR satellite loop and focus your attention on the Southwest. Since there are no clouds present, we can't really tell how the air is moving over this region. Now, check out the corresponding loop of water vapor images and focus your attention on the same area. What do you see? Do you notice the ever-so-slight counter-clockwise circulation of the air off the California coast? Such upper-level circulations are in fact important, as we will learn later in this course. The lesson learned here is that we were able to identify this circulation only with the aid of water vapor imagery.

Water vapor imagery's ability to trace upper-level winds ultimately allows forecasters to visualize upper-level winds, and computers can use water vapor imagery to approximate the entire upper-level wind field. Here's an example of such "satellite-derived winds" in the middle and upper atmosphere at 00Z on August 26, 2017 (on the far left side of the image, you can see Hurricane Harvey about to make landfall in Texas). Having such observations over the data-sparse oceans is extremely valuable to forecasters, and much of this information gets put into computer models so that they better simulate the initial state of the atmosphere, which leads to better forecasts than if we didn't have these observations.

This concludes our look at the three most common types of satellite imagery. Before moving on to radar imagery, take a moment to review the key points about water vapor imagery as well as the Case Study below.