HES Trade-Off Studies

High spectral resolution infrared radiances from the Hyperspectral Environmental Suite (HES) on Geostationary Operational Environmental Satellite (GOES-R and beyond) will allow for monitoring the evolution of atmospheric profiles and clouds. The HES is currently slated to be launched in 2013. HES, together with the Advanced Baseline Imager (ABI) will operationally provide enhanced spatial, temporal and vertical information for atmospheric soundings and clouds. Trade-off studies have been done on the spectral coverage, spectral resolution, spatial resolution, temporal resolution, band-to-band co-registration and signal-to-noise ratio. HES data applications investigated include sounding temperature/moisture retrievals, trace gas estimation, cloud retrieval and surface property retrieval. Synergistic uses of ABI and HES data for better atmospheric and cloud retrievals are also under investigation. Moderate Resolution Imaging Spectroradiometer (MODIS) and Atmospheric InfraRed Sounder (AIRS) measurements from Earth Observing System's (EOS) Aqua platform are used to demonstrate the ABI/HES system capability of deriving atmospheric, cloud and surface parameters with high accuracy.

HES-Disk Sounding (HES-DS) task

Note: click on image to enlarge.

1. HES spectral coverage study - water vapor information from IR LMW and SMW

One important issue for HES instrument design was the selection of water vapor spectral coverage. Usually IR longwave coverage (LW, approximately 650-1200cm^-1) is selected for temperature, ozone and surface property retrievals. For water vapor region, one can use either longer middlewave (LMW, approximately 1200-1650cm^-1) or shorter middlewave (SMW, approximately 1650-2250cm^-1). Selection of both water vapor sides might be a better option in terms of information. However data volume would be increased. Figure 2 shows an example of HES brightness temperature (BT) spectrum for LW (blue line), LMW (green line) and SMW (red line) (upper panel), the 14bit HES instrument noise from the Technical Requirement Document (TRD) was used in simulating the HES radiances (see the noise in the lower panel). Figure 3 shows temperature and relative humidity (RH) retrieval rmse from LW + LMW, LW + SMW, and LW + SMW with SMW noise reduced by half (NF=0.5). In general, the temperature retrieval difference between LW+LMW and LW+SMW is about 0.1K, while the water vapor retrieval difference is about 1%. With SMW noise reduced by half, both temperature and water vapor retrieval differences between LW+LMW and LW+SMW are reduced. Considering other factors for LMW (for example, lower spectral resolution than SMW, more trace gas, etc.), the temperature and moisture retrieval differences between LW + LMW and LW + SMW are very small.

Figure 1a: Spectral coverage of the current GOES sounder and two examples of preliminary HES.

Figure 1b: The current GOES sounder (lower panel) versus HES (upper panel) for water vapor mixing ratio weighting functions.

Figure 2: Examples of BT spectral for HES LW, LMW, and SMW (upper pannel). The noise in NEDR is shown in the lower pannel.

Figure 3: The temperature retrieval rmse at 1km vertical resolution and Relative Humidity (RH) retrieval rmse at 2km vertical resolution for LW + LMW, LW + SMW, and LW + SMW with SMW noise reduced by half.

2. HES spatial resolution study

The spatial resolution for HES is very important because "hole hunting" will be the effective way to find clear pixels for atmospheric sounding without microwave sounding capability on the geostationary satellite. Fine spatial resolution allows for a higher possibility of finding clear pixels. This is very important because (a) fine spatial resolution HES measurements will meet the mesoscale forecast requirement, and (b) fine spatial resolution enables one to find more homogeneous 2 by 2 or 3 by 3 fields-of-view (FOV) scenes for the possible ABI/HES cloud-clearing.

Figure 4: The 1km MODIS TPW at 1900UTC on July 20, 2002 from EOS's AQUA satellite and a number of reduced spatial resolutions for IR imager/sounders.

3. HES spectral resolution requirement for non-sounding

Trade-off studies are necessary to investigate the impact of IR LW window spectral resolution on non-sounding products. A study has been performed to demonstrate that in the IR longwave window region, a spectral resolution of 1cm-1 or better is necessary for accurately retrieving the non-sounding products such as IR surface emissivity and surface skin temperature by minimum local emissivity variance method. Figure 5 shows from upper to lower panels the BT spectrum with rock emissivity observation, true emissivity spectrum (black line), retrieved emissivity spectrum with true surface skin temperature (green line) and surface skin temperature deviated by 1K (green and red lines). The noise factor indicates the noise added in the simulation (e.g., 0.5 means half noise). The mean local emissivity variance is also indicated in each panel. The spectral resolution in the figure is 0.625 cm^-1. Figure 6 shows the emissivity variance difference between wrong skin temperature and true skin temperature as a function of skin temperature error, different lines correspond to various spectral resolutions and noise factors. It clearly indicates that a spectral resolution of 0.625 cm^-1 with half noise and nominal noise will create an accurate emissivity and skin temperature retrievals, while only half noise will create good surface property retrieval with lower spectral resolution (e.g., 1.25 cm^-1).


Figure 5: True and retrieved emissivity spectra.	Figure 6: The emissivity variance difference between wrong and true skin temperatures, as a function of skin temperature error.

4. HES spectral resolution study

High spectral resolution of HES is very important to achieve the fine structure (good vertical resolution) of atmospheric temperature and moisture profile. The vertical resolution analysis has been performed on HES example 1 (see Figure 1) with a spectral resolution of (a) 0.3, (b) 0.6, and 1.2 cm^-1, respectively.

Figure 7: HES spectral resolution impact on atmospheric temperature (upper panel) and moisture retrieval (lower panel).

5. HES signal-to-noise analysis

HES signal-to-noise is very important to achieve the good accuracy of atmospheric temperature and moisture profile. The signal-to-noise analysis has been performed on HES example 1 (see Figure 1) with a noise factor of 0.5, 1.0, and 2.0 cm^-1, respectively (TRD noise was used).

Figure 8: Signal-to-noise impact on temperature (left panel) and moisture (right panel).

Summary

(1) LMW and SMW provide similar water vapor information along with the LW spectral band.
(2) Spatial resolution is very important for clear "hole hunting" without a microwave sounder. A spatial resolution of 10km or better is required.
(3) A high spectral resolution of 1 cm^-1 or better, should be considered for the window region along with a good signal-to-noise ratio.
(4) A spectral resolution better than 1.25 cm^-1 will result in a vertical resolution better than 1km for temperature and water vapor mixing ratio.
(5) A good signal-to-noise ratio is crucial for temperature and moisture retrieval with good accuracy.

GOES-R Activities at CIMSS / SSEC