The physical quantities included in the input files are the following:
Tips and instructions on reading and using the input files
The input for each CIRC case consists of 5 or 6 files, depending on the availability of cloud and aerosol information. The list of file types is as follows:
Contains surface skin temperature, solar zenith angle, number of atmospheric levels and the broadband solar irradiance incident at TOA for the specific case.
Contains height, pressure and temperature at atmospheric levels.
Contains, pressure, temperature, and mixing ratio of atmospheric molecules in atmospheric layers.
Contains Angstrom alpha exponent, and aerosol optical depth at 1000 nm, aerosol single-scattering albedo and asymmetric factor in atmospheric layers.
Contains cloud fraction, liquid water path, ice water path, effective droplet radius and effective ice particle radius in atmospheric layers.
Contains surface albedo (assumed lambertian), the product of surface albedo and the normalized surface downwelling flux from CHARTS, and the TOA solar irradiance according to Kurucz (not case-specific, i.e., same in all files), all at 1 cm-1 resolution.
The surface emissivity is assumed to be unity for the LW calculations. Molecular concentrations are expressed as volume mixing ratios, if you need therefore to convert to mass mixing ratios, multiply by mw/28.97 where mw is the molecular weight of the particular molecule (included in layer_input_caseX.txt). To obtain aerosol optical depth at an arbitrary wavelength λ (in μm) use τ(λ)=τ(1μm)*λ-alpha . The aerosol single-scattering albedo (SSA) and asymmetry parameter (g) are spectrally invariant, and their values represent averages across the solar spectrum. Since each radiation code participating in CIRC may have a unique spectral band structure, the surface albedo is provided at relatively high spectral resolution (1 cm-1). A version of this spectral albedo that has been weighted by the CHARTS-calculated downwelling surface irradiance is also provided. You have a choice to use either the weighted or the unweighted spectral albedos, or even, to simulate perhaps common usage where downwelling spectral irradiance is unavailable for proper weighting, the Kurucz spectral solar function. When submitting results, please mention which spectral albedo you used.
Here is a sample FORTRAN code to read the contents of these files. The input files themselves can be downloaded from this page.
How the input was derived
The input quantities for the SGP and NSA cases are determined in a manner similar to that used for BBHRP. A general description of the BBHRP methodology for determining input suitable for RT calculations from observations can be found here. Some of the information therein is repeated below. Any changes applied to adapt the cases for the purposes CIRC are also documented below.
For the clear cases, the atmospheric column discretization is variable, ranging from 54 m near the surface to 4 km at the top of the atmosphere. For the cloud cases, the discretization of the cloudy layers is 62.5 m between 375 m and 1 km and 100 m above 1 km.
Profiles of temperature and humidity up to ~ 20 km are based on radiosonde information while above ~ 20 km climatological profiles are used. The exception is Case 3 for which the source is ARM's Mergedsounding product throughout the atmospheric column. The surface (skin) temperature is estimated by inverting the observed broadband LW upwelling surface irradiance, and the air temperature of the lowest atmospheric level is derived from AERI 675-680 cm-1 measurements. Each layer column amount of water vapor derived from the sonde measurement is scaled by the ratio of the total precipitable water vapor retrieved from MicroWave Radiometer (MWR) measurements and the total precipitable water vapor from the sonde. For the NSA cases, because of the high uncertainty of the MWR measurements in dry conditions, the sonde precipitable water was scaled with that retrieved from NSA extended-range AERI measurements between 535 and 560 cm-1. A CO2 mixing ratio of 360 ppmv is used for the year 2000 calculations (SGP cases; Cases 1-3, and 6), 375 ppmv is used for the 2004 NSA nominal CO2 case Case 4), 750 ppmv is used for the NSA double CO2 case Case 5), and 380 ppmv is used for the 2005 Pt. Reyes case Case 7). The ozone profiles come from monthly means from the MOZART model climatology and the total ozone column amount is scaled to agree with that measured by TOMS. The AERI-LBL comparison for Case 1 (9/25/00) improved in the 9.6 µm band when October 2000 ozone was used, we thus adopted October instead of September ozone values for this case. For all other species mixing ratios are taken from the US Standard Atmosphere.
The cloud of Case 6 is based on ARM's ARSCL product, which provides height distributions of hydrometeor reflectivity (and cloud boundaries) every 10 seconds based on observations from a Millimeter Cloud Radar (MMCR) and Micropulse Lidar (MPL). These ARSCL products are combined with thermodynamic profiles from radiosondes and column integrated water vapor estimates from the MWR and input into the Microbase cloud property retrieval, which computes a time-height grid of the liquid water concentration, liquid effective radius, ice water concentration, and ice effective radius. Within the Microbase retrieval, the initial liquid water concentration data are integrated to produce an estimate of the Liquid Water Path (LWP) and then scaled by the ratio of the LWP retrieved from coincident MWR measurements and this LWP estimate. The retrieved cloud properties for each time and height are averaged over a 20-minute interval which was empirically determined to best encompass the cloud fields affecting the irradiance measurements used for the comparisons. For Case 7 (Pt. Reyes), the cloud property retrievals are based on the MIXCRA inversion algorithm (Turner et al. 2007). Cloud top and base are determined from WACR (a W-band Doppler radar operating at 95 GHz) measurements. The cloud is assumed to be vertically homogeneous.
For the SGP clear-sky cases (Cases 1-3) the aerosol optical depths are derived from measurements of the Multi-Filter Rotating Shadowband Radiomenter (MFRSR). These MFRSR measurements, averaged over 5 minutes, are used to derive an Angstrom relationship. The aerosol single-scattering albedo is retrieved from the diffuse-to-direct ratio measured in two visible MFRSR channels. The aerosol asymmetry parameter is derived from measurements of the backscattered radiation by the surface-based Aerosol Observation System (AOS), adjusted to ambient humidity at the surface. Aerosols are assumed to occupy only the lowest (i.e., closest to the surface) six atmospheric layers. For the NSA cases (Cases 4-5) aerosol properties are obtained from the Aerosol Best (ABE) product. For all clear-sky cases, the single-scattering albedo and asymmetry factor are vertically and spectrally invariant. ABE is also the source of aerosol information for cloudy Case 6, but in this case vertical variability is allowed and aerosols extend beyond the first 6 layers. For Case 7, no aerosol measurements are available, and the aerosol loading is therefore assumed to be zero.
The SGP shortwave surface albedo values used in the calculation are based upon measurements of the upward-looking MFRSR and the downward-looking MFR, which are colocated at the 10m and 25m towers at the SGP ACRF site. For each location, the ratio of observed upward to downward irradiance in six measurement channels of the instrument are used to classify the surface type under that tower. This classification and the six ratio values are then used along with published spectral albedos to create a (piecewise continuous) spectral albedo function with 1 cm-1 resolution. The surface albedo functions associated with each tower are then averaged to obtain the specification of spectral surface albedo that is input to the reference high-resolution calculation. For NSA only a single tower is available for surface classification and albedo estimation. Based on a satellite (Multispectral Thermal Imager) image analysis it has been determined that it is appropriate to use a surface albedo estimated by taking the weighted average of the surface below the tower (85%) and open water (15%) albedos. For PYE Case 7 MODIS-derived surface (i.e. land) albedos are used to generate pseudo-MFR albedos and a spectral albedo function is then derived as described for SGP. The surface albedo is assumed to be lambertian.
For questions or problems reading/interpreting the input dataset please contact Lazaros Oreopoulos.