diff --git a/pyglider/ncprocess.py b/pyglider/ncprocess.py
index 8995298..1278b38 100644
--- a/pyglider/ncprocess.py
+++ b/pyglider/ncprocess.py
@@ -312,7 +312,7 @@ def make_gridfiles(inname, outdir, deploymentyaml, *, fnamesuffix='', dz=1, star
     # https://cfconventions.org/Data/cf-conventions/cf-conventions-1.7/build/aphs06.html
     dsout.attrs['featureType'] = 'trajectoryProfile'
     dsout['profile'].attrs['cf_role'] = 'profile_id'
-    dsout['mission_number'] = int32(1)
+    dsout['mission_number'] = np.int32(1)
     dsout['mission_number'].attrs['cf_role'] = 'trajectory_id'
     dsout = dsout.set_coords(['latitude', 'longitude', 'time'])
     for k in dsout:
diff --git a/pyglider/slocum.py b/pyglider/slocum.py
index f8fb947..c290cb1 100644
--- a/pyglider/slocum.py
+++ b/pyglider/slocum.py
@@ -789,7 +789,8 @@ def raw_to_timeseries(indir, outdir, deploymentyaml, *,
 def binary_to_timeseries(indir, cachedir, outdir, deploymentyaml, *,
                          search='*.[D|E]BD', fnamesuffix='',
                          time_base='sci_water_temp', profile_filt_time=100,
-                         profile_min_time=300, maxgap=300):
+                         profile_min_time=300, maxgap=300,
+                         replace_attrs=None):
     """
     Convert directly from binary files to netcdf timeseries file.  Requires
     dbdreader to be installed.
@@ -821,6 +822,11 @@ def binary_to_timeseries(indir, cachedir, outdir, deploymentyaml, *,
     profile_min_time : float
         minimum time to consider a profile an actual profile (seconds)
 
+    replace_attrs : dict or None
+        replace global attributes in the metadata after reading the metadata
+        file in.  Helpful when processing runs with only a couple things that
+        change.
+
     Returns
     -------
     outname : string
@@ -831,6 +837,9 @@ def binary_to_timeseries(indir, cachedir, outdir, deploymentyaml, *,
         raise ImportError('Cannot import dbdreader')
 
     deployment = utils._get_deployment(deploymentyaml)
+    if replace_attrs:
+        for att in replace_attrs:
+            deployment['metadata'][att] = replace_attrs[att]
 
     ncvar = deployment['netcdf_variables']
     device_data = deployment['glider_devices']
diff --git a/pyglider/utils.py b/pyglider/utils.py
index d3de10a..0f1b996 100644
--- a/pyglider/utils.py
+++ b/pyglider/utils.py
@@ -29,11 +29,15 @@ def get_distance_over_ground(ds):
     ds : `.xarray.Dataset`
         With ``distance_over_ground`` key.
     """
+
     good = ~np.isnan(ds.latitude + ds.longitude)
-    dist = gsw.distance(ds.longitude[good].values, ds.latitude[good].values)/1000
-    dist = np.roll(np.append(dist, 0), 1)
-    dist = np.cumsum(dist)
-    dist = np.interp(ds.time, ds.time[good], dist)
+    if np.any(good):
+        dist = gsw.distance(ds.longitude[good].values, ds.latitude[good].values)/1000
+        dist = np.roll(np.append(dist, 0), 1)
+        dist = np.cumsum(dist)
+        dist = np.interp(ds.time, ds.time[good], dist)
+    else:
+        dist = 0 * ds.latitude.values
     attr = {'long_name': 'distance over ground flown since mission start',
             'method': 'get_distance_over_ground',
             'units': 'km',
@@ -465,10 +469,17 @@ def fill_metadata(ds, metadata, sensor_data):
 
     """
     good = ~np.isnan(ds.latitude.values + ds.longitude.values)
-    ds.attrs['geospatial_lat_max'] = np.max(ds.latitude.values[good])
-    ds.attrs['geospatial_lat_min'] = np.min(ds.latitude.values[good])
-    ds.attrs['geospatial_lon_max'] = np.max(ds.longitude.values[good])
-    ds.attrs['geospatial_lon_min'] = np.min(ds.longitude.values[good])
+    if np.any(good):
+        ds.attrs['geospatial_lat_max'] = np.max(ds.latitude.values[good])
+        ds.attrs['geospatial_lat_min'] = np.min(ds.latitude.values[good])
+        ds.attrs['geospatial_lon_max'] = np.max(ds.longitude.values[good])
+        ds.attrs['geospatial_lon_min'] = np.min(ds.longitude.values[good])
+    else:
+        ds.attrs['geospatial_lat_max'] = np.nan
+        ds.attrs['geospatial_lat_min'] = np.nan
+        ds.attrs['geospatial_lon_max'] = np.nan
+        ds.attrs['geospatial_lon_min'] = np.nan
+
     ds.attrs['geospatial_lat_units'] = 'degrees_north'
     ds.attrs['geospatial_lon_units'] = 'degrees_east'
     ds.attrs['netcdf_version'] = '4.0'  # TODO get this somehow...