Made for monitoring mode:

-Update so that LiCSBAS time series are referenced correctly.
-Some small changes to outputs to terminal.

licsalert/licsalert.py

@@ -932,7 +932,7 @@ def LiCSBAS_for_LiCSAlert(LiCSAR_frame, LiCSAR_frames_dir, LiCSBAS_out_dir, logf
 
 #%%
 
-def LiCSAlert_preprocessing(displacement_r2, downsample_run=1.0, downsample_plot=0.5, verbose=True):
+def LiCSAlert_preprocessing(displacement_r2, downsample_run=1.0, downsample_plot=0.5, verbose=True, mean_centre = True):
     """A function to downsample the data at two scales (one for general working [ie to speed things up], and one 
     for faster plotting.  )  Also, data are mean centered, which is required for ICASAR and LiCSAlert.  
     Note that the downsamples are applied consecutively, so are compound (e.g. if both are 0.5, 
@@ -944,6 +944,7 @@ def LiCSAlert_preprocessing(displacement_r2, downsample_run=1.0, downsample_plot
         downsample_run | float | in range [0 1], and used to downsample the "incremental" data
         downsample_plot | float | in range [0 1] and used to downsample the data again for the "incremental_downsample" data
         verbose | boolean | if True, informatin returned to terminal
+        mean_centre | boolean | add option to control if mean centered.  
         
     Outputs:
         displacement_r2 | dict | input data stored in a dict as row vectors with a mask
@@ -956,6 +957,7 @@ def LiCSAlert_preprocessing(displacement_r2, downsample_run=1.0, downsample_plot
         2021_04_14 | MEG | Update so handle rank2 arrays of lons and lats properly.  
         2021_05_05 | MEG | Add check that lats are always the right way up, and fix bug in lons.  
         2021_10_13 | MEG | Add function to also downsample ENU grids.  
+        2021_11_15 | MEG | Add option to control mean centering, and warning that it is happening.  
     """
     import numpy as np
     from licsalert.downsample_ifgs import downsample_ifgs
@@ -967,7 +969,10 @@ def LiCSAlert_preprocessing(displacement_r2, downsample_run=1.0, downsample_plot
     shape_start = displacement_r2["mask"].shape                                                     # and the shape of the ifgs (ny, nx)
 
     # 0: Mean centre the interferograms (ie. break any connection to a reference pixel/region that the interferogram was set to)
-    displacement_r2["incremental"] = displacement_r2["incremental"] - np.mean(displacement_r2["incremental"], axis = 1)[:,np.newaxis]                            # mean centre the data (along rows) 
+    if mean_centre:
+        if verbose:
+            print(f"LiCSAlert_preprocessing: mean centering the interferograms before downsampling")
+        displacement_r2["incremental"] = displacement_r2["incremental"] - np.mean(displacement_r2["incremental"], axis = 1)[:,np.newaxis]                            # mean centre the data (along rows) 
 
     # 1: Downsample the ifgs for use in all following functions.  
     if downsample_run != 1.0:                                                                                       # if we're not actually downsampling, skip for speed

licsalert/monitoring_functions.py

@@ -216,7 +216,8 @@ def initiate_licsalert_volcano(licsalert_dir, region, volc, template_file_dir):
         with open(licsalert_dir / "all_volcs_products" / "licsalert_all_volcs_log.txt", "a") as all_volcs_log:
             message = f"A licsbas .json was found for {volc} but no licsalert directory.  The directory has now been created, and the template files copied in.  "
             print(message)
-            all_volcs_log.write(message)
+            all_volcs_log.write(message + '\n')
+
 
 
     # 1: get all the licsbas volcanoes, and loop through seeing how they compare to the files used by licsalert for that volcano
@@ -422,6 +423,13 @@ def baseline_from_names(names_list):
     licsbas_json_creation_time = licsbas_json_creation_time.replace(microsecond = 0)
 
     print(f"Opening the LiCSBAS .json file with timestamp {licsbas_json_timestamp} and system creation time {licsbas_json_creation_time}.  ")
+
+    # 0: Get the reference area.  
+    ref_list = licsbas_data['refarea'] 
+    ref_xy = {'x_start' : int(ref_list[0]),                                            # convert the correct part of the string to an integer
+              'x_stop'   : int(ref_list[1]),
+              'y_start'  : int(ref_list[2]),
+              'y_stop'   : int(ref_list[3])}
 
     # 1: get the lons and lats of each pixel in the image
     lons_mg, lats_mg = np.meshgrid(licsbas_data['x'], licsbas_data['y'])    
@@ -446,6 +454,11 @@ def baseline_from_names(names_list):
     n_im, length, width = cumulative_r3.shape                                                          # time series size, n_im = n_acquisisions
     mask_coh_water_r3 = np.repeat(mask_coh_water[np.newaxis,], n_im, axis = 0)                         # new version that has expanded to be the same size as the cumulative ifgs
 
+    # 3a: Reference the time series
+    ifg_offsets = np.nanmean(cumulative_r3[:, ref_xy['y_start']: ref_xy['y_stop'], ref_xy['x_start']: ref_xy['x_stop']], axis = (1,2))                                          # get the offset between the reference pixel/area and 0 for each time
+    cumulative_r3 = cumulative_r3 - np.repeat(np.repeat(ifg_offsets[:,np.newaxis, np.newaxis], cumulative_r3.shape[1],  axis = 1), cumulative_r3.shape[2], axis = 2)         # do the correction (first make ifg_offsets teh same size as cumulative).      
+
+    # 3b: Deal with masking etc.
     cumulative_r3_ma = ma.array(cumulative_r3, mask=mask_coh_water_r3)                                  # mask the cumulative ifgs (first one should be all 0s), note that this could still have nans in it
     for nan_pixel in np.argwhere(np.isnan(cumulative_r3_ma)):                                           # find any pixels that are nan in this, and iterate over
         mask_coh_water_r3[:, nan_pixel[1], nan_pixel[2]] = 1                                            # and modify the mask so that they are masked for all times
@@ -464,14 +477,7 @@ def baseline_from_names(names_list):
     tbaseline_info["baselines"] = baseline_from_names(tbaseline_info["ifg_dates"])                                                          # and their temporal baselines
     tbaseline_info["baselines_cumulative"] = np.cumsum(tbaseline_info["baselines"])                                                         # cumulative baslines, e.g. 12 24 36 48 etc
 
-    # 5: Get the reference area.  
-    ref_list = licsbas_data['refarea'] 
-    ref_xy = {'x_start' : int(ref_list[0]),                                            # convert the correct part of the string to an integer
-              'x_stop'   : int(ref_list[1]),
-              'y_start'  : int(ref_list[2]),
-              'y_stop'   : int(ref_list[3])}
-
-    # 6: Try to get the DEM
+    # 5: Try to get the DEM
     try:
         dem = nested_lists_to_numpy(licsbas_data['elev'])                                                 # 
         displacement_r2['dem'] = dem                                                                      # and added to the displacement dict in the same was as the lons and lats