Thermal gradients across the telescope structure can cause serious distorsions in the shape of the dish. This is particularly true during sunrise and sunset. Since normal observing at the JCMT starts before sunset, and continues after sunrise till 09:30 typically, the performance of the telescope is definitely degraded at the start of the first shift, and at the end of the second shift. We are currently implementing a system to compensate for thermally induced deformations in the telescope surface. In order to do this, we have installed 200 temperature sensors on the telescope with a system to readout each sensor in about two minutes of time, being short enough sothat the thermal gradients do not change much between measurements. A FEA model is then used to calculate the thermal expansion of the elements of the telescope structure, in particular the displacements of the nodes where the surface panels attach to the telescope. Since the surface consists of 276 panels which can all be adjusted with motorized adjusters, this provides a convenient and easy to implement a scheme to adjust the surface in almost real-time. We expect to have this active surface control system working mid 2000.

Typical temperature variations, the resulting surface distorsions and the effect on the beamshape at higher frequencies can be viewed as mpeg movies. Depending on the browser and system used, the quality of the movies can vary between reasonably good to downright poor. All three movies are made from the same temperature data set. The movie starts an hour after sunrise (around 8 am) and runs till around 5 pm.

The movie "Temperature gradients across telescope" shows a thermal image of the telescope structure, against a background color corresponding to the average temperature of the telescope. This means that the image will be faint or will completely disappear when the telescope is thermally relaxed as during night time. Large gradients across structural members of the telescope will make these elements stick out.

The second movie, "Thermally induced surface deformations" show the FEA modelled surface shape during the day. It is interesting to observe how the surface stabilizes during the day, reaching a plateau of 40-50 microns rms even when the ambient temperature inside the enclosure keeps changing. As the surface distorsions depend on temperature gradients rather than overall temperature (a uniform temperature change would only result in a trivial focus change), this shows that the gradients reach their maximum in just a few hours after sunrise, remain at the same level during the day, and disappear again in just a few hours after opening the enclosure. The plateau in the surface rms does not seem to vary very much from day to day.

The third movie, " Distorsions in the 450 micr. beamshape" shows how the beam profile at 450 microns changes during the day as a result of thermal gradients and the resulting surface node displacements.

The last movie, "False color 450 micr. beamshape" is a shorted version of the 450 microm beamshape movie in false colors. If the display on the webpages comes out very poorly, playing this version may help.