These plots show information derived from arc observations with a
slit vs. the slit width. 
</font><p><font size="2">
The change in spectral resolution is caused by using a more suitable value for one of
the arc recipe command line parameters since 2017-12. Details are given in the history
section of the plot tutorial.

</font><p><font size="2">
New 2019-02:
</font><p><font size="2">
<b>UVB</b><br>
Resolving power top level requirement (TLR) derived from design specification for UVB-arm is lam/dlam = 5000 at 1.0 arcsec slit width. 
Observed values for 2017-10...2019-02 calibrations show 5440+-14.  
Polynomial fit (slit^3) covering 2019-01 and slits 0.4x11 ... 1.6x11 shows: 5390 at 1.0''. 
Scaling the best-fit polynomial, by 5000/5390 (=0.938) results in the TLR for any slit (dashed line). 


</font><p><font size="2">
<b>VIS</b><br>
resolving power TLR for VIS arm is lam/dlam = 7500 at 1.0 arcsec slit width. 
Observed values for 1.0'' are not available.
Ploynomial fit (see Box 2) shows 7880 at 1.0 arcsec. 
Scaling the best-fit polynomial, by 7500/7880 (=0.952) results in the TLR for any slit (dashed line). 

</font><p><font size="2">
<b>UVB</b><br>
resolving power top level requirement for VIS arm is 4800 at 1.0 arcsec slit width. 
Ploynomial fit (see Box 2) shows 5120 at 1.0 arcsec.
Scaling the best-fit polynomial, by 4800/5120 (=0.938) results in the TLR for any slit (dashed line). 
 

</font><p><font size="2">
For more information see the
<a href="http://www.eso.org/observing/dfo/quality/XSHOOTER/qc/wavecal_qc1.html">plot tutorial page</a>.