Cloud heights measured by MISR from 2000 to 2015

Davies and Molloy (2012) reported a decrease in the global effective cloud height over the first 10years of Multiangle Imaging Spectroradiometer (MISR) measurements on the Terra satellite. We have reexamined their time series for possible artefacts that might especially affect the initial portion of the record when the heights appeared anomalously high. While variations in sampling were shown to be inconsequential, an artefact due to the change in equator crossing time that affected the first 2years was discovered, and this has now been corrected. That correction, together with the extension of the time series by five more years, yields no significant overall trend in global heights during the first 15years of Terra operation. The time series is dominated by large interannual fluctuations associated with La Nina events that mask any overall trend on a global scale. On a regional basis, the cloud heights showed significant interannual variations of much larger amplitude, sometimes with fairly direct cancellation between regions. There were unexplained differences between the two hemispheres in the timing of height anomalies. These differences persisted over a large range of extratropical latitudes, suggestive of teleconnections. Within the tropics, there were very strong changes associated with the Central Pacific and Indonesian Maritime Continent regions that oscillated out of phase with each other, with interannual amplitudes that exceeded 1km. Plain Language Summary Effective cloud height is an important climate variable, strongly influencing the overall greenhouse effect, with a change of approximate to 15m/decade being comparable to the rate of CO2 increase. The Multiangle Imaging Spectroradiometer (MISR) instrument on the Terra satellite uses a stereo technique to systematically measure cloud heights with a sampling accuracy of approximate to 11m on the global, annual average. Over the last 15years we find interannual departures are greatest during La Nina events, reaching over 40m on a global average. The apparent decreasing trend during the first 10years noted by Davies and Molloy has now vanished, in part due to correction for sun glint, and in part due to higher clouds in the last 5years. Regional cloud height changes continue to show a high degree of correlation, especially for about five distinct regions of equatorial clouds. Southern Hemisphere cloud heights have increased, whereas Northern Hemisphere heights have decreased.