Hyper-Spectral Enterprise Algorithm Package (HEAP)

The HEAP is an enterprise sounding data processing system that was developed by combining the Infrared Atmospheric Sounding Interferometer (IASI) data processing system and the NOAA Unique Combined Atmospheric Processing System (NUCAPS). The HEAP was developed to generate (1) spectrally and spatially thinned radiances, (2) retrieved products such as profiles of temperature, moisture, trace gases and cloud-cleared radiances, and (3) global validation products such as radiosonde matchups and gridded radiances and profiles. The thinned radiance products are produced in BUFR format using the NetCDF4 Reformatting Toolkit (N4RT) and are tailored to specifically Numerical Weather Prediction (NWP) centers. The HEAP Environmental Data Records (EDR) products are archived in Comprehensive Large Array-Data Stewardship System (CLASS) for non-real time users and can be acquired from www.nsof.class.noaa.gov.

These products are derived from the Cross-track Infrared Sounder (CrIS) and Advanced Technology Microwave Sounder (ATMS) currently onboard the Joint Polar Satellite System satellites (SNPP NOAA-20, and NOAA-21). As well as the data from the Infrared Atmospheric Sounding Interferometer (IASI), Advanced Microwave Sounding Unit-A (AMSU-A), Microwave Humidity Sounder (MHS), and Advanced Very High Resolution Radiometer onboard the European Space Agencys (ESA) Metop series of polar orbiting satellites (Metop-B/C).

HEAP Global Granules Composite Images

The HEAP global granules composite images are produced for the last seven days at the 15 fixed air pressure levels or layers. They are produced by using the NUCAPS retrievals which are derived based on a fixed air pressure variable grid: temperature is derived at the fixed pressure levels (1014 mb, 853 mb, 707 mb, 497 mb, 407 mb, 300 mb, 260 mb, 201 mb, 151 mb, 103 mb, 71.5 mb, 51.1 mb, 29.1 mb, 9.5 mb, 1.0 mb), and mixing ratio variables are derived at the layer pressure using the effective air pressure variable (1000 mb, 840 mb, 695 mb, 487 mb, 399 mb, 293 mb, 254 mb, 196 mb, 147 mb, 99.5 mb, 68.8 mb, 49.3 mb, 27.6 mb, 8.82 mb, .838 mb). Each product is computed separately for each granule, and then the global image is generated by combining the data from individual granules based on the granule geographical location. For each image the granules from the preceding 12 hours of observation are used; each image combines the granules of data measured at both ascending and descending nodes.

HEAP Global Gridded Products

Global Gridded 0.5 deg lat x 2 deg lon Images
NUCAPS EDR Global Gridded products include the Temperature (deg K), Water Vapor Mixing Ratio (g/Kg), Liquid Water Mixing Ratio (g/Kg), Ozone Mixing Ratio (ppb), Methane Mixing Ratio (ppb), Carbon Dioxide dry mixing ratio (ppm), Carbon Monoxide Mixing Ratio (ppb), Sulfur Dioxide mixing ratio (ppb), Nitric Acid Mixing Ratio (ppb), and Nitrous Oxide Mixing Ratio (ppb). The retrievals are derived based on a fixed air pressure variable grid: temperature is derived at the fixed pressure level (1014 mb, 853 mb, 707 mb, 497 mb, 407 mb, 300 mb, 260 mb, 201 mb, 151 mb, 103 mb, 71.5 mb, 51.1 mb, 29.1 mb, 9.5 mb, 1.0 mb) and mixing ratio variables are derived at the layer pressure using the effective air pressure variable (1000 mb, 840 mb, 695 mb, 487 mb, 399 mb, 293 mb, 254 mb, 196 mb, 147 mb, 99.5 mb, 68.8 mb, 49.3 mb, 27.6 mb, 8.82 mb, .838 mb).

HEAP Retrieval Statistics

The NUCAPS retrieval statistics are generated for Temperature (Tp) over two layers: average over mid-troposphere (520-790 mb) and average over full troposphere (200-1100 mb); and Water Vapor Mixing Ratio (WVMR) statistics are generated over full troposphere. The NUCAPS retrieval estimates are compared with GFS estimates to compute bias and rms error over these layers and are plotted for each granule on the 24-hour scale for the day.

To generate the temperature bias and rms error over a large ensemble of K granules one needs to take the bias for a single granule, bias(k), weighted by the number of accepted cases, Nacc(k) such as:

bias = sum {Nacc(k)*bias(k)} / sum{Nacc(k)}, where sum is for k = 1, K rms = sqrt [ {(sumNacc(k)*rms(k)^2) } / sum {Nacc(k)} ], where sum is for k = 1, K

To generate the WVMR bias and rms error over a large ensemble of K granules the following formulas are used:
bias = sum{qmean(k)*bias(k)} / sum{qmean(k)}, where sum is for k=1, K
rms = sqrt [ {sum (qmean(k)*rms(k)^2)} / sum{qmean(k)} ], where sum is for k = 1, K