Climatology and changes in cloud cover in the area of the Black, Caspian, and Aral seas (1991–2010): a comparison of surface observations with satellite and reanalysis products

Abstract

This article presents a climatology of total cloud cover (TCC) in the area of the three inland Eurasian seas (Black, Caspian, and Aral Sea). Analyses are performed on the basis of 20 years of data (1991–2010), collected from almost 200 ground stations. Average TCC is 49%, with broad spatial and seasonal variability: minimum TCC values are found in summer and to the southeast, whereas maximum values correspond to winter and to the northwest. For the whole area, linear trend analyses show that TCC did not vary during the study period. We only detected a statistically significant positive trend (+1.2% decade−1) in autumn. We obtained different results for the regions delimited by means of a principal component analysis: a clear decrease, both for the annual, spring, and summer series, was detected for the south of Black Sea, while increasing TCC was found for the annual, autumn, and winter series in the north Caucasus and the west and north of Black Sea. We also analysed the TCC data from global gridded products, including satellite projects [International Satellite Cloud Climatology Project (ISCCP), Pathfinder Atmospheres Extended (PATMOS‐x), cLoud, Albedo & Radiation (CLARA)], reanalyses [ERA‐interim, National Centers for Environmental Prediction/Department of Energy (NCEP/DOE), Modern‐Era Retrospective Analysis for Research and Applications (MERRA)], and surface observations [Climatic Research Unit (CRU)]. Although all these products capture the seasonal evolution over the study area, they differ substantially both among them and in relation to the ground observations: reanalyses produce much lower values of TCC, while ISCCP and CLARA provide a summer minimum that is too high. Trend analyses applied to these data generally showed a decrease in TCC; only CRU and NCEP/DOE tally with the ground data as regards the absence of overall trends. These results are discussed in relation to previous studies presenting trends of other variables such as sunshine duration, diurnal temperature range, or precipitation; we also discuss the connections with changes in synoptic patterns and environmental changes, in particular in the Aral Sea region.

Fewer clouds in the Mediterranean: consistency of observations and climate simulations

“…Even if all databases emphasize the decline of clouds in the Mediterranean during the last decades, the disagreement in rate and significance of the trends highlights the uncertainties of the different satellite and reanalysis products in capturing trends of clouds, in agreement with other recent studies over other regions 33 38 , even if the latter products can reproduce observed TCC trends in US 39 .…”

Decadal differences of the diurnal temperature range in the Aral Sea region at the turn of the century

“…3), from November to April the decadal difference of the DTR was negative (-0.37 C dec À1 in FMA) and positive during the rest of the year (>0.20 C dec À1 ). The negative value of the decadal DTR difference in spring is consistent with the positive trend of the total cloud cover identified in this region, which is associated with intensification of the activity of the South-Caspian, Murgab and Upper Amu Darya cyclones (Calb o et al, 2016). On the other hand, as observed in Fig.…”

Inconsistency of surface‐based (SYNOP) and satellite‐based (MODIS) cloud amount estimations due to the interpretation of cloud detection results

“…Details of how to convert the native, continuous cloud fraction (0–100%) into discrete okta values (0–8) are given in the WMO's manual on coding the results of meteorological observations [SYNOP FM‐12 code; WMO ()]. However, there are no guidelines for the reverse conversion from the okta to the continuous scale and climatologists usually assume that 1 okta equals 12.5% (Meerkötter et al, ; Matuszko, ; Sanchez‐Lorenzo and Wild, ; Calbó et al, ), and the usage of other weights is exceptionally rare (Boers et al . ()).…”