Changes to MOPITT pairing #316
Conversation
…k/MELODIES-MONET into develop_tropomifix
melodies_monet/driver.py
Outdated
| if obs.obs_type.lower() in list(self.pairing_kwargs.keys()): | ||
| pairing_kws = self.pairing_kwargs[obs.obs_type.lower()] | ||
| else: | ||
| pairing_kws = {'apply_ak': True, 'mod_to_overpass': False} |
There was a problem hiding this comment.
Need to add an error message as to why MM might break if no AK is available.
There was a problem hiding this comment.
Seems like the message could be made a bit more helpful by telling you why it's being set to true (defaulting since not set) and how you (the user) could fix it.
This comment was marked as duplicate.
This comment was marked as duplicate.
Sorry, something went wrong.
| @@ -53,7 +103,7 @@ def check_timestep(model_data,obs_data): | |||
| print('Timestep check and model resample failed') | |||
| raise | |||
|
|
|||
| def mopitt_l3_pairing(model_data,obs_data,co_ppbv_varname): | |||
| def mopitt_l3_pairing(model_data,obs_data,co_ppbv_varname,global_m=True): | |||
There was a problem hiding this comment.
Why not global_model? Only 4 more chars.
| ufsaqm: | ||
| files: '/scratch1/BMC/rcm2/rhs/monet_example/AEROMMA/UFS-AQM/cmaq54_OriRave1/aqm.202306*/12/aqm.t12z.dyn.*.nc' | ||
| mod_type: 'rrfs' | ||
| is_global: False |
There was a problem hiding this comment.
In the driver code I'm seeing isglobal, but I like this (is_global) better. May want to fix the driver.
There was a problem hiding this comment.
Also it doesn't seem to have been added to the YAML doc?
There was a problem hiding this comment.
Fixed and added a description to YAML doc.
| # Determine local time offset | ||
| local_utc_offset = np.zeros([ny,nx],dtype='timedelta64[ns]') | ||
| # pandas timedelta calculation doesn't work on ndarrays | ||
| for xi in np.arange(nx): | ||
| local_utc_offset[:,xi] = pd.to_timedelta((modobj['longitude'].isel(x=xi)/15).astype(np.int64),unit='h') | ||
|
|
||
| # initialize local time as variable | ||
| modobj['localtime'] = (['time','y','x'],np.zeros([nmt,ny,nx],dtype='datetime64[ns]')) | ||
| # fill | ||
| for ti in np.arange(nmt): | ||
| modobj['localtime'][ti] = modobj['time'][ti].data + local_utc_offset |
There was a problem hiding this comment.
You may want to round lon/15 instead of truncating by converting to int. You could try the approach here to get around the pd.to_timedelta limitation:
local_utc_offset = (modobj['longitude'] / 15).round().astype("timedelta64[h]")Then you should be able to do
modobj['localtime'] = modobj['time'] + local_utc_offset
melodies_monet/driver.py
Outdated
| if obs.obs_type.lower() in list(self.pairing_kwargs.keys()): | ||
| pairing_kws = self.pairing_kwargs[obs.obs_type.lower()] | ||
| else: | ||
| pairing_kws = {'apply_ak': True, 'mod_to_overpass': False} |
There was a problem hiding this comment.
Seems like the message could be made a bit more helpful by telling you why it's being set to true (defaulting since not set) and how you (the user) could fix it.
melodies_monet/driver.py
Outdated
| # grab kwargs for pairing. Use default if not specified | ||
| if obs.obs_type.lower() in list(self.pairing_kwargs.keys()): | ||
| pairing_kws = self.pairing_kwargs[obs.obs_type.lower()] | ||
| else: | ||
| print('WARNING: The satellite pairing option apply_ak is being set to True. Pairing will fail if there is no AK available.') | ||
| pairing_kws = {'apply_ak': True, 'mod_to_overpass': False} |
There was a problem hiding this comment.
Currently you'd have an issue later if pairing_kwargs for sat were set in the control but didn't contain apply_ak, e.g. just mod_to_overpass or another unrelated setting. An alternative approach would be to set pairing_kws to your desired defaults first and then update it based on pairing_kwargs from the control. That way you know everything is set. There would still be an opportunity to let the user know that they may need to set apply_ak.
Co-authored-by: Zachary Moon <[email protected]>
Co-authored-by: Zachary Moon <[email protected]>
Adds utility for sampling model data to satellite local overpass time.
Modifies MOPITT L3 pairing to calculate model averages using only model data at the satellite overpass time.
Also ensures that model is paired to correct obs time. Previously assumed that model and obs datasets were on same time at that point, but this isn't true if there are gaps in the obs dataset.
New argument for the yaml file if model output hasn't been pre-processed to monthly mean: mod_to_overpass.
Note: Insert/delete flagging includes some commits from other recent pull requests (#304, #312 ).