<p><strong>Abstract.</strong> It has long been assumed that Rayleigh lidar can be used to measure atmospheric temperature profiles up to about 90 or 100 km and that above this region the technique becomes invalid due to changes in atmospheric composition which affect basic assumptions on which Rayleigh lidar is based. Modern powerful Rayleigh lidars are able to measure backscatter from well above 100 km requiring a closer examination of the effects of the changing atmospheric composition on

more »

... ved Rayleigh lidar temperature profiles. <br><br> The NRLMSISE-00 model has been used to simulate lidar signal (photon-count) profiles, taking into account the effects of changing atmospheric composition, enabling a quantitative analysis of the biases and errors associated with extending Rayleigh lidar temperature measurements above 100 km. The biases associated with applying a nominal correction for the change in atmospheric composition with altitude has also been investigated. <br><br> The simulations reported here show that in practice the upper altitude limit for Rayleigh lidar is imposed more by the accuracy of the temperature or pressure used to seed the temperature retrieval algorithm than by accurate knowledge of the atmospheric composition as has long been assumed.</p>