In certain cases, HN24 can be reported erroneously. What triggers this
is e.g. a storm which deposits 5cm of low-density snow which quickly
settles. During such a storm, the HS over a 24-hour period may go from
e.g. 100cm to 105cm then back to 100cm.
The root cause was taking the difference between the highest HS value
reported over a 24-hour period and subtracting the lowest value. But
when HS fluctuates throughout a storm due to rapid settlement and/or
wind, this calculation does not provide correct data.
Additionally, this bugfix opens the door to reporting HS settlement to
InfoEx should InfoEx support that data via autostations.
remote_data[element_cd] = None
if len(hn24_values) > 0:
- hn24 = max(hn24_values) - min(hn24_values)
+ # instead of taking MAX - MIN, we want the first value (most
+ # distant) - the last value (most recent)
+ #
+ # if the result is positive, then we have HN24; if it's not,
+ # then we have settlement
+ #hn24 = max(hn24_values) - min(hn24_values)
+ hn24 = hn24_values[0] - hn24_values[len(hn24_values)-1]
+
+ if hn24 < 0.0:
+ # this case represents HS settlement
+ #
+ # TODO: determine if InfoEx supports auto-stations reporting
+ # HS settlement values
+ hn24 = 0.0
if len(wind_speed_values) > 0:
wind_speed_avg = sum(wind_speed_values) / len(wind_speed_values)