The GOES DPI Enhancement Page

Individual parameters from atmospheric profiles derived from sounding radiometers on satellites have often been displayed in an image format. At the full horizontal resolution of the satellite sounding data (nominal 10 km for GOES) and for a typical geographic coverage (no smaller than several states), an image format allows effective display of all data points (versus the plotting of numerical values). The parameters shown may be representative of either an entire profile (such as total precipitable water vapor (TPW), or lifted index (LI) stability) or just one level (such as the cloud top pressure (CTP)). As full atmospheric profiles of temperature and moisture can not be retrieved in cloudy areas with current GOES (infrared) sounders, two complimentary arrangements are typically presented with DPI: either retrieved atmospheric values in the clear regions composited with window channel (11um band-8) "brightness temperatures" for the clouds, or retrieved cloud information (such as CTP) composited with window channel depiction of the earth's surface in the clear regions.

Click here for an example of the latest TPW DPI from CIMSS. For a more extensive look at real-time GOES Sounder data, including other assorted DPI, as well as links to background material on the GOES Sounder, see The CIMSS Realtime GOES Page.     [The details on recent and current color enhancement tables for the GOES DPI indicate relatively slight, ensuing evolution over the years.]

The CIMSS DPI are stored as AREA files within McIDAS (the University of Wisconsin Man computer Interactive Data Access System). [Click here for more information on McIDAS.] The areas are navigated and calibrated. The DPI are maintained at a digital depth of 1 byte (that is, having 256 levels available (0-255)). For a profile DPI (such as TPW), that range is typically partitioned into two sections: 0-200 for parameter values and 201-255 for the "cloud" (actually window channel brightness temperatures).

An "unenhanced" DPI example is from 19 UT on 7 Oct 1998, where the data are presented as a composite of all available GOES-8 and -10 sectors within an hour, re-mapped to a 14km Mercator projection. This "absence of enhancment" simply maintains an ouptput of 0-255 equally for each color gun (blue, green, red) matching the 0-255 input (thus, at digital value 20, the "rgb" values are also each at 20; at 240, each color also at 240). Two graphical scales have been added near the bottom, just above the "gray bar" or gray wedge. The top scale show the 0-255 range; the bottom, the DPI value scale.

A few values, notably at digital levels of 60, 120, 180, and 240, have been identified in the next example, by coloring those individual values (in blue, yellow, red, and orange respectively). The calibration for this DPI example of the TPW is simple (within the 0-200 portion). TPW is one third of the digital count (TPW=D/3). Thus, those digital levels of 60, 120, and 180 correspond to TPW values of 20, 40, and 60mm. Note the convention show here, that the right edge of the units numeral on the graphical scale lines up with the enhanced level highlighted in the gray bar. For the cloud portion, the input values of 201-255 correspond to temperatures of 301-193K. Levels of 0C (freezing point) and -32/-52C (used as thresholds in areal definition of mesoscale convective complexes) were arbitrarily chosen for later enhancement of that portion.

A not uncommon ratio of clear to cloudy regions (crudely 50/50) is obvious in these unenhanced examples. One variation, to de-emphasize the cloudiness, is to simple stretch the available gray shades over just the clear portion, setting the clouds to black. Of course, the addition of color to the display does indeed help to portray the pattern of the moisture field. An example of this employs the "standard" TPW enhancement table (ET) used at CIMSS. However, the remainder of this page will emphasize an assortment of other ETs, either already in use at other sites or newly suggested.

Currently, three enhancement tables (ET) are initially shown here for total precipitable water (TPW). For each ET, a batch file is available, for listing and/or copying, which replicates the EU command sequence on McIDAS. The tables are simply referred to by number (10, 20... as in PW010, PW020...). Following each table are examples of recent re-mapped mercator composites of GOES-8/10 TPW DPI shown with that ET.

• Enhancement PW010 specifications.
  1. For 09 Jul 1998 .
  2. For 29 Jul 1998 .
  3. For 24 Aug 1998 .
  4. For 09 Sep 1998 .
  5. For 07 Oct 1998 .
  6. For 20 Oct 1998 .
  7. For 06 Jan 1999 .
• Enhancement PW020 specifications.
  1. For 09 Jul 1998 .
  2. For 29 Jul 1998 .
  3. For 24 Aug 1998 .
  4. For 09 Sep 1998 .
  5. For 07 Oct 1998 .
  6. For 20 Oct 1998 .
  7. For 06 Jan 1999 .
• Enhancement PW030 specifications.
  1. For 09 Jul 1998 .
  2. For 29 Jul 1998 .
  3. For 24 Aug 1998 .
  4. For 09 Sep 1998 .
  5. For 07 Oct 1998 .
  6. For 20 Oct 1998 .
  7. For 06 Jan 1999 .

Below find a few subsections of the large domain coverage with all of the sample enhancements. The sectors are smaller, allowing for simultaneous viewing of the different tables.

The use in an enhancement, of those similar color pairs which color-challenged people have difficulty distinguishing (green/yellow, magenta/blue, white/cyan) presents an ineffective display (for them). By continuity and/or gradient considerations, they may be able to correctly associate an enhancement level with a given color in an image; but, usually it is not obviously evident. More limited use of different colors generally presents a less confusing display.

For example, look at the samples shown below. How well do the patterns and information content appear in the images shown on the right side ?

The above images, both at 1300 UT on 24 Aug (top) and 09 Sep (bottom) 1998, show a side-by-side comparison of PW010.ET (left) versus the equivalent (right) of how that might appear to a color-challenged individual. To note a few specific problematic locations on the 24 Aug (top) example, look at: "lighter yellow-green" in both W cen TX Pan (20mm) and ext NW AR (29mm); "darker yellow-green" in both N cen MI (24mm) and cen SW QUE (33mm); and, "moderate green-orange" in both SE WY (12mm) and SE cen NE (39mm). On the 09 Sep (bottom) example, see: "orange-green" in both ext NW SD (19mm) and cen ND (32mm).

Similar to the previous set of examples but for a different color scheme, the above examples, both at 1300 UT on 24 Aug (top) and 09 Sep (bottom) 1998, show a side-by-side comparison of PW020.ET (left) versus the equivalent (right) of how that might appear to a color-challenged individual. To note a few specific problematic locations on the 24 Aug (top) example, look at: "yellow-green" in both cen IN (50mm) and NW NE (16mm). On the 09 Sep (bootom) example, see: "yellow-green" in both E VA (14mm) and E cen TX (48mm).

The new proposed "consensus" enhancement shown above is an attempt to use attractive attributes from the previously discussed examples, illustrated for a wide range of different patterns (from moist summer to dry winter conditions).

This new "consensus" enhancement, to be used as the first default for TPW DPI, is available below, provided as a McIDAS batch file. Its current name is simply: GDPW1.ET. Displays of the default enhancement are also provided below, matching the large scale, seasonal variation displays of the initial enhancements shown near the beginning of this page. The projections are re-mapped mercator composites of both GOES-8 and -10 DPI.

• Enhancement GDPW1 specifications.
  1. For 09 Jul 1998 .
  2. For 29 Jul 1998 .
  3. For 24 Aug 1998 .
  4. For 09 Sep 1998 .
  5. For 07 Oct 1998 .
  6. For 20 Oct 1998 .
  7. For 06 Jan 1999 .

Current GOES DPI Sites

RAMSDIS Online at CIRA/CSU

NWS Western Region Headquarters (at SLC UT)

NOAA Forecast Products Development Team (FPDT) at ORA (Wash DC)

NOAA Advanced Satellite Products Team (ASPT) at UW-CIMSS