We report on a statistical study of prominence eruptions (PEs) using a catalog of these events routinely imaged by the Atmospheric Imaging Assembly (AIA) on board the Solar Dynamics Observatory (SDO) in the 304 r{A} pass band. Using an algorithm developed as part of an LWS project, we have detected PEs in 304 r{A} synoptic images with 2-min cadence since May 2010. A catalog of these PEs is made available online (https://cdaw.gsfc.nasa.gov/CME_list/autope/). The 304 r{A} images are polar-transformed and divided by a background map (pixels with minimum intensity during one day) to get the ratio maps above the limb. The prominence regions are defined as pixels with a ratio $ge$2. Two prominence regions with more than 50% of pixels overlapping are considered the same prominence. If the height of a prominence increases monotonically in 5 successive images, it is considered eruptive. All the PEs seen above the limb are detected by the routine, but only PEs with width $ge$15{deg} are included in the catalog to eliminate polar jets and other small-scale mass motions. The identifications are also cross-checked with the PEs identified in Nobeyama Radioheliograph images (http://solar.nro.nao.ac.jp/norh/html/prominence/). The catalog gives the date, time, central position angle, latitude, and width of the eruptive prominence. The catalog also provides links to JavaScript movies that combine SDO/AIA images with GOES soft X-ray data to identify the associated flares, and with SOHO/LASCO C2 images to identify the associated coronal mass ejections. We examined the statistical properties of the PEs and found that the high-latitude PE speed decreased with the decreasing of the average polar magnetic field strength of the previous cycle.
Download