A daytime over land algorithm for computing AVHRR convective cloud climatologies for the Iberian Peninsula and the Balearic Islands

Abstract

ABSTRACT:A daytime over land multispectral cloud detection algorithm is presented to derive accurate convective cloud climatologies with high spatial resolution (1.1 km) over the Iberian Peninsula (IP) and the Balearic Islands (BI). The cloud detection scheme was designed to process Advanced Very High Resolution Radiometer (AVHRR) HRPT data and is tested here on NOAA-17 morning (0900-1200 UTC) and NOAA-16 afternoon (1200-1500 UTC) overpasses for the warm 6-month study period May-October. The algorithm consists of four spectral threshold tests applied to each pixel. Test 1 corresponds to the snow-ice removal, test 2 is the thermal infrared test, test 3 is the albedo or visible test and test 4 is the ratio between near-infrared and visible channels. The algorithm discretizes all AVHRR data into four groups called cloud-filled, cloud-free, snow-ice and snow-free radiances. The high-resolution convective cloud masks are obtained by subtracting snow-ice pixels from cloudy ones. In this article, a detailed description of the convective cloud detection scheme and the sources of error detected for each test are given, and the first seasonal and monthly regional convective cloud frequency composites are presented. Future applications of the newly proposed threshold algorithm in climate and meteorology are also discussed in this article, particularly the production of convective cloud composites for climate monitoring of storms over the IP and BI.

AVHRR warm‐season cloud climatologies under various synoptic regimes across the Iberian Peninsula and the Balearic Islands

“…forecasters. Apart from the marked latitudinal gradient found in cloudiness with the maximum (73.8%) over the northernmost fringe of the Atlantic‐Cantabrian area and the Pyrenees, partly attributed to low‐stratiform clouds linked to large‐scale northwesterly winds and the influence of cold fronts (Azorin‐Molina et al ., ), and the minimum in cloudiness (12.4%) over the southern region of the IP caused by the suppression of convection under the influence of the Azores high‐pressure system; we identified six ROIs as those displaying high cloud frequency amounts. These ROIs basically correspond to the main mountainous areas of the IP and the BI: (1) the Cantabrian Mountains (ROIs1; maximum cloud amount 66.2%), (2) the Pyrenees (ROIs2; 67.6%), (3) the Central System (ROIs3; 41.5%), (4) the eastern Iberian System Mountains (ROIs4; 44.5%), (5) the Betic Mountains (ROIs5; 43.9%) and (6) the centre of the isle of Mallorca (ROIs6; 41.2%) as a result of the low‐level convergence of sea breezes in the centre of the isle (Ramis and Alonso, ).…”

“…63.4% and min. 22.7%) in accordance with the occurrence of cold fronts that represents an instability approaching from the Atlantic Ocean (Azorin‐Molina et al ., ), and the development of mesoscale convective systems over the Mediterranean area.…”

“…A detailed description of the daytime over land cloud‐masking algorithm is found in the study by Azorin‐Molina et al . ().…”

“…Azorin‐Molina et al . () employed the term ‘convective’ cloud frequency composites because the algorithm was specifically defined for identifying convection and most clouds developed within the mid‐afternoon overpasses; these clouds are mostly associated with convective processes during the warm‐season of May to October. However, it should be noted that composites presented in this study not only include cloud masks that comprise larger Cumulus and Cumulonimbus clouds, but also contain convection embedded in multilevel cloud cover, and mid‐afternoon clouds linked to other synoptic‐scale situations (e.g.…”

“…frontal passages). A detailed description of the ROIs1 CSW N NW CN CNW ASE AS SE AE CS ROIs2 AN N ANE NE CNE ASW CSE S AS ASE ROIs3 CSW CW W SW CNW AE ASE ANE A AN ROIs4 CSW CNE CN CW CE ASW ASE A AE SE ROIs5 CW CNE CSW SW S ASW ANW AN A ANE ROIs6 CW CN CSW C CNE ASW CS ASE CSE ANW daytime over land cloud-masking algorithm is found in the study by Azorin-Molina et al (2013).…”