Cloud observations are performed routinely for synoptic and aeronautical purposes (WMO, 1996; ICAO, 2004). The cloud observations are traditionally performed by human observers. Several cloud layers can be reported and for each layer the cloud amount, height and type are estimated. While cloud cover and particularly cloud type are determined visually, the determination of the cloud base height can be enhanced by using radio soundings, tethered balloons or balloons with a known ascent rate, and/or searchlights. More recently, ceilometers are employed that provide the observer with frequent readings, typically every 15 seconds, of the cloud base height directly overhead. Presently, several countries, e.g. Sweden, USA and the Netherlands, perform automated cloud observations using a ceilometer in combination with a cloud algorithm that transforms the ceilometer cloud base readings in a certain time interval into cloud layers with corresponding amount and height. The automated cloud observations have been compared to visual observations (e.g. Ramsay and Nadolski, 1998, Perez et al., 2002; Wauben et al., 2006). The overall agreement is generally good, but as could be expected, situations with large differences between visual and automated cloud observations do occur. The main reason for these differences is the lack of spatial representativeness of the ceilometer measurements.
A study showed that the combination of the results of 3 instead of 1 ceilometer in the cloud algorithm did not significantly improve the overall results (Wauben, 2002; Ravilla et al., 2002). Observing systems that can provide spatial cloud information are e.g. radiometers on board satellites, camera systems and scanning systems. The space-borne instruments provide useful cloud information particularly for high clouds, but has limitations for low clouds and experiences difficulties with partially clouded pixels and semi-transparent situations. In addition satellite instruments do not give information on the cloud base height. Camera systems are currently often used as remote observing systems (e.g. Mammen and Wienert, 2005), although systems are available that automatically evaluate the images and provide an automated total cloud cover (e.g. Long et al., 2006). However, visual camera systems only give useful information during daytime and twilight and they do not give information on the cloud base height, although stereoscopy using 2 wide-angled cameras makes it possible to obtain cloud base height (and wind) information (cf. e.g. Seiz et al., 2002). Infrared camera systems require regular maintenance and are currently too expensive to be considered for operational use (Keogh et al., 2000). Several scanning systems are available. Scanning ceilometer systems are expensive and since a ceilometer measurement requires an integration time, scanning the entire sky is too time consuming. Scanning infrared radiometers (pyrometers) are less expensive and make nearly instantaneous measurements. Furthermore scanning infrared (IR) radiometers can be operated during day- and nighttime and provide, through the observed temperature, information on the height of the observed clouds. Furthermore the observed cloud base temperature itself is a useful quantity that can be combined with the cloud top temperature obtained from satellites. Hence scanning pyrometers seem promising observation systems that can give useful information on clouds.
Currently 2 types of scanning IR radiometers are commercially available. The first one is the Cloud Infrared Radiometer (CIR) manufactured by Atmos Sarl. This instrument (see e.g. Genkova et al., 2004), consists of up to 13 pyrometers attached to an arc to cover the various elevations. The arc rotates 360 degrees in order to measure the whole sky. A measurement in 30 azimuth directions takes about 3 minutes. The field of view of each IR radiometer is 6 degrees. The other scanning IR radiometer is the so-called Nubiscope manufactured by IMK/Sattler-SES (http://Sattler-SES.de/Nubiscope-US.html). This sensor consists of a single IR radiometer mounted on a pan and tilt unit. A measurement of the whole sky, every 3 degrees in elevation and every 10 degrees in azimuth takes about 6 minutes. The field of view of the IR radiometer is 3 degrees. Both scanning IR radiometers have been tested by users and give promising results. The advantages of the Nubiscope are the smaller FOV, the employment of a single radiometer and the statement that the system is weather proof and requires little maintenance. Therefore the Nubiscope was selected for a test.
In this report no detailed evaluation of the above mentioned observing systems for spatial cloud observations will be performed. The evaluation will be restricted to the Nubiscope. During a brief field test the technical aspects of the system will be investigated e.g. reliability, robustness, weather proof, required maintenance, sensitivity to contamination. A comparison of the Nubiscope results with the automated cloud observations using a ceilometer gives an indication of the quality of the results and the added value of the Nubiscope.
Wiel Wauben. Evaluation of the Nubiscope
KNMI number: TR-291, Year: 2006, Pages: 46