Site Map





If the SOLAR2000 Research Grade V2.23 GUI application is run with a number of days to model that extends past the last valid date the program will ungraciously stop. Why?

The GUI is being asked to display invalid dates and/or indices and the program must be restarted. In this case, simply exit IDL, start a new IDL session, and restart the GUI, taking care to not run the model past the last valid date.


If the "Print" button on the GUI is used for printing the plot file the right-hand border of the displayed plot may not be reproduced. Is there another solution for this?

The file "s2k_plot.tif" is the default plot file and is fully displayed with graphic software. Other print options include using the cursor to click and highlight one of the three plot panels then print that panel from IDL. Another alternative is to highlight a panel, copy the plot panel contents into the computer's internal buffer or 'clipboard', then paste, print, and/or save the graphic within a graphics program.


3. The September 1, 2000 release of SOLAR2000 RG did not properly display the irradiances in the output file for NT platforms. Can you fix this?

The display was fixed in the September 25, 2000 release (1.05a).


The combined selection of "energy flux" and "1 nm" in the GUI does not produce a tabular listing of energy flux on a 1 nm grid. Only the indices are displayed. How do I produce this combination?

The combination was provided in the February 10, 2001 release (1.15).


The program takes considerably longer to run on a DEC alpha workstation. The symptom is that the user clicks. the "Run" button and an hourglass cursor icon comes up on the GUI. window. After some time, the proper files are created and the session run is completed. Is this normal?

As of March 9, 2001 the reason for this is not yet understood and is being investigated. One possible source of latency could be a large number of files in the directory that the GUI is located within. However, there may be IDL- DEC compatibility issues that are not yet understood.


On some Unix systems (DEC alpha), clicking the "Print" button brings up a printer manager window that does not easily allow printing of the s2k_plot.tif file with Unix- based commands such as" lp -P <printer> s2k_plot.tif". How can I print the tif files?

One can manually print the" s2k_plot.tif" file that is located in the SOLAR2000 directory directly from a separate window using the appropriate Unix printer command ("lp -P <printer> s2k_plot.tif").


IDL, in this GUI application, does not properly display colors if the monitor is set to 16-bit color. Why?

Set the monitor to 8-bit or, as an alternative, 24-bit color.


I would like to have the capability of using a callable IDL subroutine module rather than a GUI. This would serve better in an IDL or Fortran routine that needs solar irradiances updated in by a time step. Can you provide this?

This capability is being developed. It will be possible to use this as a special subroutine in the future.


I would like to have the model extended to shorter wavelengths and higher energies in the X-rays (0.05 - 1.8 nm). When will this be available?

This capability is planned for a future upgrade of SOLAR2000.


There is a discontinuity for E81 values between the beginning of one time frame run and the end of a previous run. How do I remove this?

This artifact was minimized in v1.16 by using an 81-day convolution. However, a user is advised to create a data set of desired length and to use his or her own smooth algorithm.


Some of the highest resolution wavelengths in the 809 template are not uniform in grid-spacing and are repeated. Is this correct?

The SC21REFW format is used for the 809 wavelength bin format and this format does not have uniform spacing between the lines. In addition, some values may appear to be repeated, e.g., 6.63 nm, but is a result of truncation of finer resolution lines (6.630 Fe XII and 6.637 Fe XVI). The v1.21 upgrade changed the lines to be reported with 3 decimal places to avoid this truncation confusion.


How do I find the uncertainty in the irradiances and how is it calculated?

The 1 sigma error of reported instrument measurement error plus linear regression uncertainty can be found as a fraction (1.0 = 100%) in the 809 wavelength (highest resolution) listing. In the other listings, there are placeholders right now. However, since they are all coarser resolution than the 809 listing, a simple wavelength bin- average would be acceptable. A future version will incorporate the 1 sigma error for all formats.


13. Is it possible to get irradiances closer to the current epoch, nowcast, or forecast irradiances in addition to the Research grade version?

Yes, there are system, operational, and professional grade SOLAR2000 versions that produce all these irradiances. For more information, including high time resolution data, please contact SpaceWx or visit the products and services link at Additionally, NOAA/SEC hosts the daily nowcast SOLAR2000 irradiances on their server at


Can I retrieve SOLAR2000 products without using IDL?

E10.7 and Lyman-alpha are available for 5 solar cycles at the products links at These are daily historical data for Research use. Current epoch, daily irradiance products are also provided at that web site location. In the second half of 2003, SET plans to release historical 3-hourly data sets as well as forecast data sets, both of which will be provided for a fee. It may be possible to run SOLAR2000 from a web browser in the future and we are Researching this capability.


Why doesn't the S(t) value of the SOLAR2000 output match the variability in the total solar irradiance (TSI)?

The solar integrated spectrum (often referred to as S(t) in SOLAR2000 literature) is only variable between 1-122 nm in versions 1.yz. It is the integrated spectrum and will in- creasingly become the same as the variable TSI in versions 2.yz, 3.yz, and 4.yz. In version 2.yz, the spectrum is variable between 1-420 nm but does not include irradiance deficit due to sunspots.


16. Why are there double entries for some of the wavelengths longwards of Lyman-alpha?

The ASTM-E490 placeholder spectrum was used in earlier versions and had a problem of double entries. That placeholder spectrum was upgraded to the officially released version of the reference spectrum starting with SOLAR2000 version 1.24.


Why is the value of E10.7 in older model versions different from the v2.21 release?

As the accuracy and precision of the EUV wavelengths are improved in Model releases, the value of E10.7 will change since it is the inte-grated 1-105 nm EUV energy flux and the latter changes. As measurements improve through time, the value of E10.7, though changing, will more and more approach a fixed level for given solar conditions. For example, the 1-sigma E10 v1.15:v1.24 ratio is 6.15% while v1.20:v1.24 is 0.38%. See the discussion in version 2.21 release related to "Known differences..." items 3 and 4. We have identified a process to create a " standard" by which to judge E10.7 and all other integrated irradiance proxies and work is progressing to create this standard. Our AFRL colleagues will create a several solar cycle derived F10.7 from an ensemble of satellite drag data and this independent measure of what the general level of solar flux should have been will be used to evaluate future E10.7.


Is there a Fortran version of SOLAR2000 available?

A Fortran version of the model is not under development since most of our efforts are going into improvement in various spectral ranges. However, current and selected historical data files are available at the web site. Also, the web site points to links with older EUV models that are written in Fortran.


Are the plotted values correct in the v2.21 GUI window, row 2, panel (e) (EUV spectrum)?

No, the units are high by a factor of 1E3. A typo in the v2.21 plot module uses the energy flux rather than SI units as in row 2, panel (d). This has been corrected in v1.24a, v1.25, and versions after v2.21.


Which exospheric temperatures are being shown in the Tinf column of the S2K_output.txt file?

In all versions except for v2.21 these are the daytime average exospheric temperature. They are not the subsolar point temperature as has been reported earlier since a cos(72) (72 deg is the angle between local atmosphere normal vector and the Sun-line vector) is used in the physics-based 1DTD model. This angle was selected because the 1DTD temp- eratures and neutral densities for 1976-200 (SC21REFW) and 1979-050 were very close to the MSIS-86 output for those dates. In addition, this angle provides an average daytime exospheric temperature for 6 years (Jan 1 1996 - Dec 31 2001) that is 1.3 times the J70 nighttime minimum exospheric temperature for the same period. The 1-sigma variation about this average temperature is +/-6.4%. Altogether, these point to a self-consistent exospheric temperature for input spectral solar flux. 1DTD does not use the E10.7 integrated irradiance proxy but uses the 39 wavelength groups and lines that are common in aeronomy. In v2.21, a conversion was applied to the temperatures to produce an absolute value very close to J70 nighttime minimum exospheric temperatures. We have decided not to use this conversion in all other releases of this Tinf parameter.


Is there a SC21REFW type of template for wavelengths between 121.5 and 199.5 nm?

We plan to extend the current SC21REFW type of wavelength temp- late into this spectral region in an upcoming release of the model.


22. Why are there slight differences in the output numbers between the RG and PG model versions?

The S2K PG proxy file is generated on an operational server and will have slightly different values for either the F10 or Ly-a and/or their 81-day smoothed values. This will translate to small output differences in the integrated irradiance proxies since the data on the server is continually being updated and may also be running with a different version of S2K.


Why do v2.21 and 2.23 have the same results for the integrated irradiance proxies and some wavelength formats for models B and C?

Model C uses the VUV2002 model and it's results are currently pro- vided only through the 1 nm spectral resolution formats. A future upgrade will include VUV2002 values through all formats and will change the integrated irradiance proxies to reflect calculations with that model.