Peering at Precipitation (Extras) (Copy)

Explore Further...

Key Data Resources

Perhaps the best resource on the Web for accessing products from active and passive microwave sensors aboard satellites is the Naval Research Laboratory's (NRL) Tropical Cyclone page (opens in a new window). On this site, you can view a wide array of satellite data for current tropical cyclones, or you can select a past storm from their "Storm Library."

After you've selected a storm of interest, along the left side of the page, you'll see many options for "Platforms," "Sensors," and "Products." For the record, many of the products covered in the previous section come from passive microwave sensors such as these:

• Weather System Follow-on Program Microwave Imager (opens in a new window) (WMI)
• Advanced Microwave Scanning Radiometer 3 (opens in a new window) (AMSR3)
• GPM Microwave Imager (GMI) 
• IMERG (which isn't really a sensor itself as you just learned, but it's listed along with them on this site). 

Under "Products," you'll see many options that mostly take the form of a combination of letters and numbers. The numbers typically represent the frequency used to create the product (for example, products with "91" refer to 91-GHz imagery). Clicking on one of these products will show you the available imagery of that type for your chosen storm. But, in the discussion of 85-91-GHz imagery, I mentioned that a handful of "twists" on standard 85-91-GHz imagery exist, and that's where the letters come in. The standard images that you learned about are listed by the frequency used followed by the letter "H" (for "horizontal polarization") on the NRL page. But, one of the drawbacks of such images is that the brightness temperatures in ocean areas with few clouds (away from tropical cyclones) can be relatively low (the ocean doesn't emit much microwave radiation). Sometimes, they're low enough that they can be similar to those found in modest areas of deep convection within a tropical cyclone, which could get confusing (since the precipitation in each area is likely much different). 

To correct this issue, the NRL page has a product that uses something called "polarization-corrected temperatures" (the products have "PCT" in the name) which effectively eliminates the possible confusion with the ocean or low cloud areas and focuses on precipitation in the layer between roughly five and nine kilometers. To see an example of the difference, check out the image slideshow below. The first image is a standard "91H" image of a tropical depression. Notice the green shading toward the top left of the image, which represents brightness temperatures similar to some areas of convection within the tropical depression. Now, advance the image slideshow to see the corresponding "91PCT" image, which is superimposed upon traditional IR imagery. The structure of the deep convection within the tropical depression really stands out with this product.

For weaker tropical cyclones (tropical depressions or tropical storms), the products labeled "HW" may also be of interest (the "W" actually stands for "weak"). In a nutshell, these products use a different color scheme to spotlight higher brightness temperatures, which are more consistent with the "relatively modest" convection in tropical storms and tropical depressions (less attenuation by sparser concentrations of precipitation-sized ice particles). As a result, the microwave footprints of tropical storms and tropical depressions are easier to observe on this special imagery. 

You may also notice products using other frequencies and with other letters attached. If you're interested in finding out the "twists" of these products, I encourage you to do some digging on your own, but I wanted to highlight a few of the more prominent variations on 85-91-GHz imagery. The NRL site also has images that show quantitative precipitation estimates from IMERG, but you may also be interested in some of the products available on the GPM site. Their data visualization site (opens in a new window) includes several options for viewing real-time and archived data. Remember that IMERG combines all available passive microwave rain data from GPM and other polar orbiting satellites and adjusts per GPM calibration and even rain gauge data where available. In this way, meteorologists try to minimize the weaknesses and capitalize on the strengths of the various IR and microwave estimates that are currently available from space.

Locating the center: 85-91-GHz Imagery vs. 36-37-GHz

One of the important uses of 85-91-GHz and 36-37-GHz imagery is that these microwave images can help forecasters see the core structure of a tropical cyclone even when it's masked by high clouds on conventional satellite imagery. Being able to see the "hidden eyes" of tropical cyclones also helps forecasters pinpoint the center of a tropical cyclone when it's outside the range of aircraft reconnaissance. But, as I mentioned before, 36-37-GHz imagery is a better choice than 85-91-GHz imagery for locating a tropical cyclone's center. Let's explore the reason more in-depth.

For starters, you'll learn later that that eye-wall thunderstorms tend to lean outward with increasing altitude. To see what I mean, check out this schematic displaying a vertical cross section through a hurricane (opens in a new window). In light of this "stadium effect" and the fact that passive microwave sensors sample high altitudes within eye-wall thunderstorms (and outer rain-band storms) with 85-91-GHz radiation, it stands to reason that the diameter of the eye on 85-91-GHz images tends to be larger than the diameter at low altitudes. For example, the image slideshow below shows an 89-GHz image of Super Typhoon Bavi, which was on the verge of Category 5 intensity at the time. The image shows a fairly large eye at high altitudes in the storm. If you advance the image slideshow, you'll see the corresponding 37-GHz image, which shows that the eye had a notably smaller diameter at lower altitudes in the storm.

If you look closely, you might also notice that the location of the center of the eye on each image isn't quite the same. Does that mean the eye is tilted? Not necessarily. It turns out that an inherent error associated with the viewing geometry of the satellite exists, which is more significant in the context of 85-91-GHz imagery because the passive microwave sensor samples relatively high altitudes within eye-wall thunderstorms. Check out this schematic (opens in a new window) (not drawn to scale), which illustrates the problem that arises from the viewing geometry of the satellite. Focus your attention on a point above the freezing level in an eye-wall thunderstorm. This point lies directly above Point X (on the earth's surface). The passive microwave sensor onboard the satellite detects 89-GHz radiation upwelling from this point. But, given the angled view of the satellite, the source of this radiation, relative to the earth's surface, appears to be located at Point Y. Satellite meteorologists refer to this displacement (the satellite-perceived offset from Point X to Point Y) as parallax error.

Because of the relatively large parallax error, professional meteorologists don't usually look at the eye of a hurricane on 85-91-GHz imagery to estimate the center of circulation. Instead, they utilize 36-37-GHz imagery because the parallax error is smaller (opens in a new window) since the source of the radiation reaching the satellite comes from lower altitudes. Furthermore, the smaller, circular eye on 36-37-GHz imagery reduces the potential error while trying to locate the center of circulation (compared to 85-91-GHz imagery). Thus, 36-37 GHz imagery gives forecasters a more accurate way to determine the center of circulation of a tropical cyclone over remote seas.