Wiki Display Wiki Display

GOES-R Cloud Optical Depth

Frequently Asked Questions about the GOES-R Cloud Optical Depth Product

 

1) What is this product?

The GOES-R Cloud Optical Depth product calculates the optical depth of clouds.  It is produced for all pixels designated as "cloudy" by the GOES-R Cloud Mask algorithm. The resolution of this product varies depending on the type of image you are looking at.  For CONUS images, the resolution is 2km.  For Full Disk images, the resolution is 4km.  The GOES-R Cloud Optical Depth product is created during both day and night, however the daytime algorithm and the nighttime algorithm are complete different.  Aerosol optical depth is a unit-less quantity, with a range from 1 - 50 during the day and 1 - 8 at night.  The product's accuracy depends upon the phase of the cloud (liquid or ice) and also the time of day.  For liquid phase clouds during both the day and night, it is accurate to within 20% of the true cloud optical depth.  For ice phase clouds during the day, it is accurate to within 20% of the true cloud optical depth.  For ice phase clouds during at night, it is accurate to within 30% of the true cloud optical depth.  The GOES-R Cloud Optical Depth product is considered quantitative out to 65 degrees local zenith angle and qualitative beyond that.1

This product is useful when trying to identify the optical thickness of cloud in the atmosphere.  The higher the optical depth, the thicker the cloud.  This can inform parameters such as strength of convection, visibility, and even rain rate.

 

2) How often do I receive this data?

The cadence of the Cloud Optical Depth product is the same regardless of scan mode.  That is, the product is produced every 15 minutes for both Full Disk and CONUS images.

 

3) How do I display this product in AWIPS-II?

To display this product in AWIPS-II, go to the "GOES-R" tab of the CAVE menu, then select "Derived Products."  From there, select the region of interest (GOES-E, GOES-W, or GOES Test and Full Disk or CONUS).  Then, select the Cloud Optical Depth product.

Alternately, use the AWIPS Product Browser.  Select "Sat", then either "GOES-16" or "GOES-17".  From there, choose "Full Disk" or "CONUS", then select "COD."

 

4) How do I interpret the color maps associated with this product?

Color tables have not yet been created for this product.

 

5) What other imagery/products might I use in conjunction with this product?

This product can be used in conjunction with visible 0.64um satellite imagery to identify optically thick clouds.  In visible satellite imagery, the brighter a cloud looks, the more optically thick a cloud is.  Optical thickness is directly proportional to the moisture density of a cloud, and also the vertical depth of a cloud.  The more moisture (cloud droplets) present in a cloud, the longer the path length of a photon, therefore causing a larger cloud optical depth measurement.  If the cloud is very large, this also contributes to an increased photon path length.

By knowing which clouds contain more moisture, one can infer higher rain rates associated with that cloud.  The same is true for vertically deep clouds.

At night, this product is useful because of the lack of visible satellite imagery.  Without this, it is much more difficult to differentiate between deep convective clouds and high clouds.

 

6) How is this product created?

During the day, this algorithm utilizes the 0.64 um and the 2.2 um channels.  At night, it utilizes the 3.9 um, 11.2 um, and 12.3 um channels

 

Cloud optical depth and cloud particle size distribution describe almost completely the radiative properties of a cloud. They characterize the impact of clouds on the energy and radiative budget of the Earth, which is why both properties are  used to parameterize clouds in global climate models. Precise retrievals are critical to improving climate models.

To briefly describe the underlying idea of the retrieval, cloud optical depth, referred to absorption-free wavelengths (for instance to 550 nm), is determine by  the amount of light scattering by loud droplets. The size of the droplets is  responsible for absorption and the transition to a new direction of scattered photons, expressed by the phase function P(ζ), which is a function of thescattering angle,  ζ. Since REF is a measure of the volume of cloud particles, it is  mirrored in absorption amount of clouds. The  basic premise is that COD and CPS are inferred from solving the radiative  transfer equation for a single-layered, plan-parallel homogeneously distributed cloud above a Lambertian surface. The  retrieval  concept is based on a 1-D radiation concept where a cloud completely covers a pixel.
 
The  daytime algorithm uses an absorption-free channel to retrieve the cloud optical depth by measuring upward backscattered radiance and uses an absorption solar channel to estimate particle size through the observed amount of  absorption. Simultaneous measurements are required since estimating optical depth from backscattered signals requires the  phase  function. The  amount of absorption  cannot be separated from extinction by scattering without measurements in a
conservative channel, such as the ABI channel 2.

An adequate transformation scheme is established to transform the radiance measurements into reflectivity quantities by considering the geometrical constellation. A doubling/adding radiative transfer model (RTM) is used to solve the forward problem, i.e., the derivation of satellite sensor signals (radiance) by simulating the transfer of solar radiation through  the atmosphere for given cloud parameters. Additionally,  the RTM calculates transmittance and spherical albedo of a cloud layer.  Inferring the optical properties from measured satellite radiances is called the inverse problem. This problem will be managed by a 1D-var optimal estimation approach.  A priori assumptions and covariance matrices depend on prior knowledge of climate data sets.

 

The retrieval strategy includes in general:

•Applying a radiative transfer model to quantify the influence of the cloud microphysical parameters on the backscattered solar radiation measured at the sensor.  

•Generating look-up-tables (LUT) for cloud reflectivity of one channel in visible spectrum at 0.6 μm and for one near-infrared channel at 2.2 μm  for a wide range of possible sun/sensor geometry constellations.

•Receiving from the processing framework all other GOES-ABI derived (cloud mask, cloud height, and cloud phase) and ancillary data needed by algorithm.

•Using 1D-var optimal estimation inversion techniques to retrieve the optical thickness from LUTs of channel reflectivity based on optimal estimation method.

 

The ancillary data used in the Daytime Cloud Optical Depth product includes:

- Land Mask: Land-sea mask

- Surface Reflectance: MODIS white sky albedo

- Snow Mask: IMS derived snow mask

- Pressure Profile: NWP pressure profile with "nlev" number of levels

- Temperature Profile: NWP temperature profile with "nlev" number of levels

- Water Vapor Profiles: NCEP water vapor profile data with "nlev" number of levels

- Ozone Amount: NCEP ozone amount data

- Surface Pressure: NCEP surface height data

- Cloud Properties: Cloud reflectance as a function of effective radius, optical depth, solar zenith angle, local zenith angle, and relative azimuth difference angle.

- Transmission Coefficients LUT: Three ozone transmission coefficients for channel 2, three water vapor transmission coefficients for channels 2 and 6.

 

The derived data used in the Daytime Cloud Optical Depth product includes:

- GOES-R Cloud Mask: A cloud mask is required to determine which pixels are cloudy and which are cloud free, which in turn determines which pixels are processed.

- GOES-R Cloud Top Pressure: Cloud top pressure is required to determine the amount of absorber mass by water vapor above the cloud for atmospheric correction.

- GOES-R Cloud Phase: Cloud phase is required to determine which LUT, ice or water are used for forward model calculations.

- GOES-R Snow Cover: Using the snow mask, each pixel is flagged internally as snow or clear. In addition, if a pixel has a 11 μm brightness temperature of greater than 277K, the snow mask is turned off.

 

The ancillary data used in the Nighttime Cloud Optical Depth product includes:

- Land Mask/ Surface Type: A global land cover classification collection created by The University of Maryland Department of Geography. Imagery from  the AVHRR satellites acquired between 1981 and 1994 were used to distinguish fourteen land cover classes (http://glcf.umiacs.umd.edu/data/landcover/). This product is available at 1 km pixel resolution.

- Surface Emissivity of Channels 7, 14, and 15:  A global database of monthly infrared land surface emissivity derived using input from the Moderate Resolution Imaging  Spectroradiometer (MODIS) operational land surface emissivity product (MOD11). Emissivity is available globally at ten
wavelengths  (3.6, 4.3, 5.0, 5.8, 7.6, 8.3, 9.3, 10.8, 12.1, and 14.3 microns) with 0.05 degree spatial resolution. The  monthly emissivities have been integrated over the ABI spectral response functions to match the ABI channels.

- Clear-Sky Infrared Radiative Transfer Model Calculations: Clear-sky top-of-atmosphere (TOA) radiances and brightness temperatures computed for channels 7, 14 and 15. Profiles of  clear-sky transmission and radiance are required for the same channels, as well as the surface temperatures. Currently, these clear-sky temperatures and radiances, as well as the radiance and transmission profiles, are obtained by using a  fast clear-sky Radiative Transfer 17 Model (RTM), the Pressure-layer Fast algorithm for Atmospheric Transmisstance
(PFAST) with 101 vertical levels that match the temperature profiles described below in the All-sky Temperature profile explanation.

- All-Sky Temperature Profiles: Knowledge of the atmospheric temperature profiles is required in order to place cloud temperatures at the appropriate  level. Currently, these profiles are from GFS, the Numerical Weather Prediction (NWP) data sets available in the offline framework. These profiles are temporally interpolated from 6- or 12-hour model data, horizontally interpolated to either   0.5° or 1.0° grids and vertically interpolated to 101 levels.

- Calibration Coefficients: Due to lack of accurate calibration in some SEVIRI channels and the possibility that some ABI channels will need refined calibration during NCOMP processing, the capability to read and utilize instrument-specific calibration coefficients is included. In the 100% delivery only SEVIRI channel-7 brightness temperatures require recalibration, so in that circumstance  a calibration is applied. Future versions can have similar calibration procedures for additional channels, but only SEVIRI currently has active recalibration as it is the ABI proxy dataset. A simple slope and offset formulation is used  and the file contains a description of the calibration source. These coefficients are read from the same text file as the cloud emittance parameterization coefficients because those are also instrument-specific.

- Cloud Emittance Parameterization Coefficients: The retrieval uses a set of coefficients that allows the invocation of a parameterization that computes cloud effective  emittances for a set of 16 cloud particle size models, both water and ice, as a function of local zenith angle, clear-sky temperature, and cloud temperature for each  of the 3 ABI channels  currently used. These parameterizations,have been calculated for a fixed set of 8 cloud optical depth bins and the resultant cloud emittances are used in the algorithm  for  computing cloud temperatures in channels 7, 14 and 15 for each pixel. For a given channel, 30 coefficients are contained  in the file, hence the file contains 240 coefficients per channel for each of the 7 water droplet and 9 ice crystal models, i.e., 3840 coefficients per channel. The coefficients are in theory instrument-specific, but the same set of coefficients can usually be used for instruments with similar spectral responses in a given channel. These coefficients are read from the same text file as the calibration coefficients because those are also instrument-specific.

 

The derived data used in the Nighttime Cloud Optical Depth product includes:

- ABI Cloud Type:  Cloud type and phase are derived prior to the invocation of the NCOMP algorithm. Currently, rather than using the ABI  Cloud Phase, the values for ABI Cloud Type are input to the NCOMP algorithm where phase is then determined internally by  combining various cloud types. The ABI phase product is determined in a similar manner, but the Nighttime algorithm is  currently using its own internal combination scheme. Neither the ABI cloud phase or cloud mask products are being used directly because ABI cloud type results provide additional information and retain flexibility for future enhancements of the Nighttime algorithm.  The Nighttime algorithm results are not impacted by this internal combination scheme; it serves  only to facilitate potential future enhancements. In addition, the internally produced cloud phase allows for processing flags to be set if NCOMP or the ABI Cloud Type product provides an indication that the phase might be ambiguous, e.g., for mixed, multi-layered or super-cooled cloud types. This will enhance validation studies.

- ABI Cloud Top Temperature: Cloud top temperature is derived prior to the invocation of the Nighttime algorithm.1

 

1Walther, Andi, William Straka, and Andrew Heidinger: NOAA NESDIS Center for Satellite Applications and Research (ABI) Algorithm Theoretical Basis Document for Daytime Cloud Optical and Microphysical Properties (DCOMP) v.2.0. 6 June 2011. http://www.goes-r.gov/products/ATBDs/baseline/Cloud_DCOMP_v2.0_no_color.pdf

2Minnis, Patrick, and Patrick Heck: NOAA NESDIS Center for Satellite Applications and Research GOES-R Advanced Basline Imager (ABI) Algorithm Theoretical Basis Document for Nighttime Cloud Optical Depth, Cloud Particle Size, Cloud Ice Water Path, and Cloud Liquid Water Path v.2.0. 15 July 2010. http://www.goes-r.gov/products/ATBDs/baseline/Cloud_NCOMP_2%200_no_color.pdf

0 Attachments
2300 Views
Average (0 Votes)
Comments