On the quantification of atmospheric rivers precipitation from space: Composite assessments and case studies over the eastern north pacific ocean and the Western United States

Atmospheric rivers (ARs) are often associated with extreme precipitation, which can lead to flooding or alleviate droughts. A decade (2003-12) of landfalling ARs impacting the North American west coast (between 32.5° and 52.5°N) is collected to assess the skill of five commonly used satellite-based precipitation products [T3B42, T3B42 real-time (T3B42RT), CPC morphing technique (CMORPH), PERSIANN, and PERSIANN-Cloud Classification System (CCS)] in capturing ARs' precipitation rate and pattern. AR detection was carried out using a database containing twice-daily satellite-based integrated water vapor composite observations. It was found that satellite products are more consistent over ocean than land and often significantly underestimate precipitation rate over land compared to ground observations. Incorrect detection of precipitation from IR-based methods is prevalent over snow and ice surfaces where microwave estimates often show underestimation or missing data. Bias adjustment using ground observation is found very effective to improve satellite products, but it also raises concern regarding near-real-time applicability of satellite products for ARs. The analysis using individual case studies (6-8 January and 13-14 October 2009) and an ensemble of AR events suggests that further advancement in capturing orographic precipitation and precipitation over cold and frozen surfaces is needed to more reliably quantify AR precipitation from space.