The latest package does more in less time, and is approximately ten times faster than the previous operational version 2.79. The package includes:
- he executable SHIC-software (version 3_075)
- a new user friendly interface/user-shell for personal computers (SHICCALL version 5.25)
- a user guide.
The user-interface SHICcall allows:
- the rapid installation of new instruments
- an easy adjustment of the main settings in the program
- the evaluation of large data sets in one run.
Furthermore, the user could restrict or modify the output of the program, and the main results can be viewed immediately:
- spectral wavelength shifts
- spectral distortions (spikes)
- spectral transmission/variability during the scan
- diurnal changes in wavelength shifts
- diurnal spectral shape errors
- diurnal transmission
Also the main flagging parameters can be easily viewed from a viewer program. Flagged spectra are grouped according to the flags allowing the easy identification of problematic and erroneous spectra in large data sets. Summary of main improvements of the SHIC methodology:
- reduction of calculation time (present version (3_075) is more than ten times faster than previous release (version 2.79)); typically the full-calculation time is now 0.4-0.8 sec for each spectrum from 285-500 nm (on an 1800 MHz P4 pc)
- analysis of wavelengths up to 600 nm is possible
- extension of the viewing capabilities of the program: diurnal
changes in shifts and shape-errors, view of spectral shifts and
spectral shape errors
flagging analysis and further quality analysis on lowest reliable irradiance levels recorded and spectral shape-errors; where possible flagging is in line with WMO-guide lines for spectrometer requirements.
- calculation of spectral transmission (measurement/ET-spectrum)
- estimate of cloud-transmission effects, and cloud variability during the scan
- calculation of significance of determined wavelength shifts, which takes into account the slit characteristics of the instrument
- Ability to zip results obtained directly from the user shell (daily zip-files)
Uncertainties of the SHIC-techniques in relation to the slit characteristics have been extensively tested, and the methods used to apply the quality-flagging have been checked with alternative approaches, where appropriate and available:
- The accuracy of the determination of the wavelength shift/error in SHIC has been studied by a comparison with a well characterised spectral instrument, and wavelength errors were found to agree within 0.005 nm with lamp measurements (results obtained by Groebner, presented at EGS 2001). \The significance of the wavelength shift determination in relation to the slit characteristics (slit width) is now calculated in the analysis. Shifts are accepted as reliable if the measured and modelled local spectral structures have a significant match in local shape at 1 % significance level. In addition shifts are ignored if the shape (error ratio) and the calculated uncertainty in the shift exceed certain limits.
- The identification of the lowest wavelength and lowest
irradiance level at which the measured spectrum matches the
expected (local) shape, has been compared with results obtained
using the ratio method from Bernhard et al. An overall good
agreement was found. Flags related to the lowest irradiance level
are chosen in line with the WMO-recommendations.
The SHIC-identification of spikes in spectra has been extensively compared with a method developed at LAP (Meleti and Bais) using over 6000 spectra from LAP with occasional spikes. The results show very good agreement between both methods.
- The analysis of potential irradiance scale-errors in the UVA will be compared further with the FAST-RT model from NPI (Engelsen). Preliminary results show the results are comparable. Black flagging levels for extremely high irradiance levels: twofold higher than the expected clear sky values and black flagging for irradiance levels below 0.01 times the clear sky level are now implemented.
The SHICrivm package is developed by dr H Slaper from RIVMNational Institute for Public Health and the Environment
within the context of the EU-projects SUVDAMA and EDUCE.