What Is The MLI correction?

In roughly October 2017, the thermal blanketing on the rear of the FPMA telescope suffered a tear. This resulted in an enhanced soft X-ray throughput for FPMA compared with FPMB after this occurred. The NuSTAR calibration team updated the effective area of the optics to compensate for the loss of X-ray attenuation in the MLI in the NuSTAR 20200429 CALDB release (and the subsequent CALDB and NuSTARDAS releases). The MLI correction accounts for the additional degradation in the MLI as a function of time and in most cases requires no action from the user, as NuSTARDAS will automatically select the appropriate ARF for the observation date.

Extensive documentation of this issue may be found in Madsen, K.K., et al. (2020)   arXiv:2005.00569

Temperature Dependence of the MLI Correction between 2017 and 2020

There is evidence for a small number of short periods of time between 2017 and 2020 in which the optics became colder, indicating a further, temporary loss of MLI coverage. For observations in these periods, the model of the MLI behavior may not fully track the performance of the FPMA optic. Systematic errors are expected to be on the order of a few percent, so this is primarily visible for high signal-to-noise observations (here meaning bright X-ray binaries with high count rates) as a deviation between FPMA and FPMB at low (E<10 keV) energies.

To account for these temperature variations seen between 2017 and 2020, a temperature-dependent MLI correction was introduced in the August 2020 NuSTARDAS v2.0.0 release (with HEASoft 6.28) that used a particular thermistor to determine whether additional MLI correction was required. The number of observations where the temperature-dependent correction is required is very small, so to avoid confusion the default behavior of nuproducts was changed in the July 2021 NuSTARDAS v2.1.1 release to not apply the temperature dependent correction. As of late 2022, the thermistor used to drive this temperature-dependent correction is no longer providing accurate thermal readings. The temperature-dependent correction should not be used after December 2022 for any observation. An update to NuSTARDAS is in progress that will prevent the user from accidentally applying the temperature-dependent correction outside of the intended time range.

What if I'm still seeing enhanced FPMA emission?

We do note that there may be a subset of bright sources for which the MLI correction is not accurate. This especially true after 2022, when we no longer have accurate temperature information to determine whether an additional MLI correction is required before inspecting the data. For these observations, we provide an Xspec mtable model that may be used to apply an additional MLI correction on top of that already applied in the ARF. Care must be taken in applying this mtable as it may potentially skew the results if not used properly. When in doubt, please contact the NuSTAR SOC for guidance on the use of this mtable.

This mtable model should not be used to force a model into agreement, but only if significant differences are observed between FPMA and FPMB at low energies (E<10 keV).

The table below lists the observation sequenceIDs where the MLI fraction is below 0.9 and where we recommend the use of the mtable model (or the temperature-dependent correction).

How Is It Used?

The MLI mtable correction is made available for pathological cases when there is a substantial difference between FPMA and FPMB at low energies. If these steps are not followed and the mtable parameters and the constants between A and B are allowed to vary freely, incorrect estimates of the continuum and line shapes may occur.

1. First, ignore FPMA between 3 and 7 keV when doing your original fits. Make sure the model includes a constant offset between FPMA and FPMB. Find the value of the constant and freeze it.

2. Notice FPMA between 3 and 7 keV and add the mtable to the model. Set the value of the MLI fraction to 1.0 for FPMB and freeze it. Allow only the MLI fraction of FPMA to fit.

3. Please note: The range of MLI fraction should now lie between 0.9 and 1.0, and if the range is outside that, something else may be wrong.

Note: In some instances, the MLI correction may be too aggressive. This may be seen as the flux of FPMA being lower than FPMB between 3-5 keV, and the reason for this is that the change in MLI is gradual whereas the corrections are a step function.
In these cases one can revert to the old FPMA ARF. This can be done by running the keyword   inarffile=   in nuproducts and point it to   caldb/data/nustar/fpm/bcf/arf/nuA20100101v006.arf

Available MLI models:

nuMLIv1.mod


    Generated May 01, 2020

NuSTAR project website:   nustar.caltech.edu

• Home
• 2: QA Operations
• 5: Links