The effective horizontal range is usually 25,000 ft., but...
Ceilometers measure cloud directly above the sensor. At 10,000 ft. above the sensor the LIDAR beam is only about 40 ft. wide.
The algorithm uses past data, usually 15-30 minutes and determines the cloud height and amount. The most recent data is weighed more.
If the cloud that passes directly over the sensor isn't representative, then the sky condition reported isn't accurate.
Also, if the base of the cloud is above 5,000 ft it is usually not reported (depending on country).
A network of sensors is more accurate, but apparently "not worth the cost" in most situations.
https://www.google.com/url?sa=t&sour...Uq-ff6TDpp0r9I
https://www.google.com/url?sa=t&sour...0hqmnaBDneaSIs
Most modern ceilometers measure to 25,000 ft., but data this high is rarely used in aviation reports.
TCU and CB can be automated and derived from ceilometer, radar and satellite data, but these automated reports in my experience are unreliable.
From the links above (ICAO references):
When an observer is not available, ceilometer(s) provide the base information for clouds.
If airports have only single ceilometer (with algorithm), the cloud reporting capability is rather restricted.
In typical case the single ceilometer reports cloudiness for the line from centre of the aerodrome to
30-minutes downwind. The report totally lacks cloud data in upwind direction. The “area” covered with
the cloudiness report is highly affected with wind speed. In 30 minutes clouds are moving downwind
from the centre of the aerodrome: 2 km, with 1 m/s, 9 km with 5 m/s, 16 km with 9 m/s and 36 km with
20 m/s. The last distance is far out from the aerodrome vicinity area regardless of the aerodrome size.
Multi-ceilometer algorithms provide an enhancement for sky condition information.
Studies have shown that single and multi-ceilometer algorithms provide similar results even over 90 per
cent of cases. One of the reasons is the use of a 30-minute period that smoothens the differences. Also a
significant portion of steady conditions make this, at least partially, a statistical “fact”. This statistical
point of view discards the safety and efficiency point of view where rapid change situations have a much
greater effect on the airport operations than steady conditions. As METAR reports are not meant for
operational decisions, a single ceilometer with sky condition algorithm seems to be a reasonable
compromise.
Automatic systems with a single ceilometer and 30-minute algorithms typically suffer
with slow and late reporting of cloudiness in changing weather conditions. One possible benefit multiple
ceilometers wide apart can provide is that the calculation period for the algorithm can be shortened. The
same number of measurement points can be collected in less time reducing the wind speed effect to the
measurement area. For example, three ceilometers and a 10-minute calculation period provide the same
number of measurement data, but can better represent the aerodrome area and adapt faster to changing
weather (cloudiness) conditions.