The spatial and temporal effect of cloud cover on the acquisition of high quality landsat imagery in the European Arctic sector

Abstract

Cloud cover is one of several constraints on the ability of VIS / NIR sensors, such as those on board Landsat, to provide a continuing time series of data for glaciological studies in polar regions; for example, the long term observation of annual mass balance fluctuations or the short-term monitoring of surging glaciers. In order to ascertain the extent to which cloudiness prevents Landsat imagery from providing such datasets in the European Arctic, cloud cover data for all imagery acquired during April–September, 1983–1992, for 15 glaciologically significant areas (each represented by a Landsat path / row) were examined. Cloudiness values were derived from the Eurimage LEDA (onLine Earthnet Data Availability) catalogue. The results indicate that regions of the European Arctic may be classified into three groups on the basis of the frequency of availability of high quality imagery ( < 35% cloud cover). Marked seasonal differences are also evident between the groups. Comparable findings from previous cloud cover studies using ground station data illustrate that the scenes used in this study are representative of the regions in which they lie. The relative abilities of Landsat MSS / TM and ERS-1 SAR data to provide multitemporal datasets are compared; although SAR is unaffected by cloud, ERS-1 has a longer revisit interval (multidisciplinary phase) than Landsat. Only in East Greenland has Landsat generated a sufficient number of high quality scenes to provide a useful multitemporal dataset. Elsewhere in the European Arctic the prevalence of the cloud cover is such that it may prevent the effective use of VIS / NIR data for certain glaciological studies. If the required information can be derived from radar data, the utilization of SAR should prove advantageous for obtaining future multitemporal datasets in these regions of the European Arctic sector.