bug fix

when pairing was performed for multiple model datasets scaling of mopitt observation occured each time. This carried over after pairing for each model because observation dataset was being shallow copied into the paired dataset.

melodies_monet/util/satellite_utilities.py

@@ -102,7 +102,7 @@ def mopitt_l3_pairing(model_data,obs_data,co_ppbv_varname):
 
     # Add variable name to smoothed model dataarray, combine with obs_data
     smoothed = smoothed.rename(co_ppbv_varname+'_column_model')
-    ds = xr.merge([smoothed,obs_data]) 
+    ds = xr.merge([smoothed,obs_data.copy(deep=True)]) 
 
     # Apply scaling to drop scientific notation (x10^{18} molec/cm2 instead of molec/cm2)
     ##  Taylor plot doesn't work if don't do this.