资源环境科技发展态势分析平台

A total of 14 years (September 2002 to September 2016) of Aqua Moderate Resolution Imaging Spectro-radiometer (MODIS) monthly mean cloud data are used to quantify possible changes in the cloud vertical distribution over the tropical Atlantic.

For the analysis multiple linear regression techniques are used. For the investigated time period significant linear changes were found in the domain-averaged cloud-top height (CTH) (-178m per decade), the high-cloud fraction (HCF) (-0.0006 per decade), and the low-cloud amount (0.001 per decade). The interannual variability of the time series (especially CTH and HCF) is highly influenced by the El NinoSouthern Oscillation (ENSO). Separating the time series into two phases, we quantified the linear change associated with the transition from more La Nina-like conditions to a phase with El Nino conditions (Phase 2) and vice versa (Phase 1). The transition from negative to positive ENSO conditions was related to a decrease in total cloud fraction (TCF) (-0.018 per decade; not significant) due to a reduction in the high-cloud amount (0 :024 per decade; significant). Observed anomalies in the mean CTH were found to be mainly caused by changes in HCF rather than by anomalies in the height of cloud tops themselves.

Using the large-scale vertical motion omega at 500 hPa (from ERA-Interim ECMWF reanalysis data), the observed anomalies were linked to ENSO-induced changes in the atmospheric large-scale dynamics. The most significant and largest changes were found in regions with strong large-scale upward movements near the Equator.

Despite the fact that with passive imagers such as MODIS it is not possible to vertically resolve clouds, this study shows the great potential for large-scale analysis of possible changes in the cloud vertical distribution due to the changing climate by using vertically resolved cloud cover and linking those changes to large-scale dynamics using other observations or model data.