Open howardisaacson opened 9 months ago
Here are the 294 ThArs calculated with a single WLS, all orderlets are the same. The smoothly varying slope is desirable but the scatter within a single night is still larger than we would like.
Rather than use another single WLS for all orders, as stated previously, I will next try to use Aaron's recently produced WLS's to re-run this ThAr set. If the results are similar to those from SoCal, then the RMS should be very small. The curved slope should be removed and the RMS should be around a trend with no slope.
Here is the current develop branch (equivalent) using the second order wavelength solutions that Aaron posted. The RMS is pretty consistent. Maybe excluding the first two orders would produce better RVs.
Using the dataset described in this issue, which has 294 ThAr from 24Oct2023 to 09Nov2023, I will run the following tests. For some of the tests, I will start with only 10 ThAr observations, from the two sets of data taken on UT 5Nov2023.
Test 1: Use a single WLS for all observations and all orderlets (1,2,3,cal,sky) as the starting guess.
Test2: Does using the cal fiber as a calibration improve RV RMS?
Baseline is 5.8 m/s over 294 RVs. It uses a masters file for each individual set of calibrations (2x per day) for the WLS initial guess. These are standard pipeline RVs.