Feature Identification Exercises
Print Version
Introduction and Overview

Identifying clouds over snow on the ground is an important and difficult problem for the meteorologist concerned with short-term forecasts impacting both aviation and ground operations. For example, low stratus and fog affect surface visibilities and may indicate the presence of icing.

Orographic clouds may provide clues to the presence of mountain wave turbulence.
There are a variety of different cloud features that help the forecaster determine a range of atmospheric conditions that may impact an analysis or forecast.

Discriminating between low clouds, snow, and ice cover poses many challenges when analyzing imagery at conventional visible and longwave infrared window wavelengths. Snow, ice, and clouds all reflect incoming solar energy and thus take on a similar appearance at visible wavelengths.

At longwave infrared wavelengths where snow, ice, bare ground, and low clouds often radiate at similar temperatures, especially at night, distinguishing one feature from another becomes a difficult task.

This set of mini-examples demonstrates a multi-channel strategy to address the snow, ice, and low cloud identification problem. Visible and longwave infrared imagery will be examined in combination with imagery from two shortwave infrared channels.

A shortwave infrared imaging capability is available with the GOES, NOAA, and NASA-EOS (Terra and Aqua) satellites. A similar capability is planned for NPOESS and METOP. Centered at either 1.6 or between 3.5 and 4 micrometers, these channels detect varying amounts of reflected solar radiation depending on whether water is in liquid or ice form, and whether it is present in a cloud or on the surface.

More 1.6-micrometer imagery is becoming available as new satellites are launched. As of early 2003, satellites with a 1.6-micrometer channel include NOAA-16, -17, Terra, and Aqua.

In the absence of animation, a broader multispectral analysis offers a significantly enhanced approach to detecting clouds, snow, and ice cover. The key to this approach is shortwave infrared imagery between 3.5 and 4 micrometers and a relatively new 1.6-micrometer channel capability available with NOAA-AVHRR and Terra and Aqua MODIS. In combination with visible and infrared imagery, shortwave infrared imagery can help distinguish clouds from surface types and determine cloud phase, as well as differences between snow and ice cover. Water clouds between 3.5 and 4 micrometers are more reflective and appear brighter (warmer) than ice clouds, snow, and ice cover. A 1.6-micrometer channel has the added value of being able to distinguish between ground features and most cloud types.

Low clouds, snow, and ice often appear indistinguishable in visible and longwave infrared imagery, especially in the absence of animation.

Visible Channel

Advantages

The best uses for the visible channel for feature detection include:

· Clouds, snow, and ice appear relatively bright (reflective)
  compared to bare ground and open water

· This channel is best used in combination with shortwave infrared
  channel (1.6 or 3.5 to 4 micrometers) for distinguishing clouds
  from snow and ice cover

The limitations for feature detection using the visible channel include:

· Distinguishing between clouds, snow, ice, and other
  reflective surfaces, particularly in the absence of animation

· Detecting open water when surrounded by dense forest
  because both are poorly reflective in the visible

Channel 6 SWIR (1.6-micrometer)

Advantages

The best uses of the 1.6-micrometer shortwave IR channel for feature detection include:

· This channel excels at distinguishing clouds and cloud boundaries,
  especially from snow and ice covered surfaces. Water clouds
  between 3.5 and 4 micrometers are more reflective and appear
  brighter (warmer) than ice clouds, snow, and ice cover.

· This channel was recently added as the daytime shortwave
  infrared channel on NOAA-16 and -17 (replaced with a
  3.7-micrometer channel during nighttime) and is a dedicated
  channel available via direct broadcast with the MODIS instrument
  on board NASA-EOS's Terra and Aqua polar orbiters

· In combination with the visible channel, it can distinguish
  clouds from snow and ice cover

Limitations

The limitations of using the using 1.6-micrometer shortwave IR channel for feature detection include:

· It can be difficult to distinguish between clouds and bare ground
  since bare ground tends to be more reflective at 1.6 micrometers

· It is more difficult to discriminate between snow and ice cover
  compared to viewing with channels between 3.5 and 4 micrometers

Channels between 3.5 and 4 micrometers

***NOTE: In this module, we represent channel range using
MODIS channel 22, centered at 3.9 micrometers. ***

Advantages

The best uses of channels in the 3.5- to 4-micrometer range for feature detection include:

· Distinguishing water clouds (highly reflective and warm)
  from ice cloud, snow and ice cover (poorly reflective and cold),
  and colder bare ground

· Detecting open water when surrounding ground is significantly
  colder or warmer, and not covered by snow and/or ice. Open water
  and vegetated surfaces, particularly forests composed of needle
  trees, are poorly reflective and difficult to distinguish in the visible.

· Detecting water clouds at night when used in combination with
  the longwave infrared channel

· Detecting ice clouds in combination with visible and 1.6-micrometer
  imagery when available, or visible and longwave infrared channels

· Distinguishing low clouds from snow and ice cover in combination
  with the visible channel

· Discriminating high clouds from snow and ice cover in combination
  with longwave infrared channel

Limitations

The limitations of using the using the 3.5- to 4-micrometer range shortwave IR channel for feature detection include:

· Difficulty distinguishing ice clouds from snow and ice cover as both
  are poorly reflective and appear relatively dark or cold
  in the imagery

· Difficulty distinguishing water clouds from open water or warm land.
  A water cloud reflects incoming solar energy such that its
  brightness temperature can approach the temperature of
  adjacent water or bare ground

Channel 31 (11-micrometer) IR Window

Advantages

The best uses of the 11-micrometer IR window imagery for feature detection include:

· Detecting relatively cold cloud-top features associated with
   convection, cirriform, and mid-level cloud types

· Can be used in conjunction with cloud phase information obtained
  from shortwave infrared channels to help determine potential
  for icing

· Can be used in combination with shortwave infrared imagery for
  discriminating ice clouds at night. During daytime, it can
  be used to discriminate high clouds from snow and ice cover or in
  combination with shortwave IR and visible imagery.

· Can be used in conjunction with a shortwave infrared channel
   (between 3.5 and 4 micrometers) to detect water clouds at night

Limitations

The limitations of the 11-micrometer IR window channel for feature detection include:

· Distinguishing low clouds from relatively cold ground or water,
   a condition common during wintertime and at night

· Distinguishing between snow, ice cover, and adjacent frozen
   ground is very difficult

· Alone it cannot confirm the presence of supercooled water clouds,
  generally present between 0 and -15 degrees C