QC on GPS data + removing dlr/ulr for bad t_rad + recalculating RH from averaged vapour pressures

[BaptisteVandecrux]

Persistence QC is done on both raw and tx data, and extended to rh and wspd

https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/process/L1toL2.py#L69-L73 https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/qc/persistence.py#L23-L24

QC on GPS data (#238)

Done through

• a persistence QC on gps_lat, gps_lon

• a persistence QC on gps_alt https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/qc/persistence.py#L20-L21 If variable 'gps_lat_lon' is given then only times when both gps_lat and gps_lon are static are being removed https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/qc/persistence.py#L79-L100

• removing all the elevations that deviate from an elevation baseline (gap-filled monthly median) by more than 100 m

• removing gps_lat and gps_lon when gps_alt is missing or has been removed by the steps above https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/process/L1toL2.py#L75-L81

See for example here. All the pink points have missing or bad gps_alt. The baseline elevation is the black dashed line.

Removing dlr and ulr when t_rad is bad or not available (#240)

• static t_rad are identified and removed using persistence QC https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/qc/persistence.py#L22
• once QC has been done (either manually or with persistence_qc), dlr and ulr are filtered from non-null t_rad https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/process/L1toL2.py#L83-L86

Re-calculating rh from time-average vapour pressures (#193)

RH is defined as the ratio between partial vapour pressure and saturation vapour pressure. When calculating hourly, daily, monthly averages, the ratio of the mean should be taken instead of the mean of the ratio (currently used)

https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/process/aws.py#L749-L759

https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/process/aws.py#L767-L805

removing percentile QC

see #242

smoothing and gap-filling tilt (#176, #123)

This is necessary for the calculation of dsr_cor and usr_cor. It is done there: https://github.com/GEUS-Glaciology-and-Climate/pypromice/blob/304a2706954fb7b2521901807477a3f33e94656e/src/pypromice/process/L1toL2.py#L121-L142

[ladsmund]

Next time: Please split up the updates into multiple pull requests. 🙏 This will make it much easier to read and discussion the changes 😅