An objective scheme for deriving cloud parameters from ARM observations and evaluating the performance of operational NWP models.
| O'Connor, Ewan | University of Reading |
| Gaussiat, Nicolas | University of Reading |
| Hogan, Robin | University of Reading |
| Illingworth, Anthony | University of Reading |
Operational NWP models represent clouds by one or two prognostic variables, such as cloud fraction and mean water content in each grid box. Vertical profiles of cloud radar and lidar backscatter together with microwave radiometer brightness temperatures can provide data to evaluate the performance of such models in representing clouds, but NWP modellers find it difficult to interpret the raw observations. Based on the EU 'CloudNET' project we have developed a scheme which firstly categorises the returns in terms of a target, such as liquid, ice or mixed phase clouds, drizzle or precipitation, aerosols, insects or ground clutter. Based on this objective categorisation, a series of algorithms are called which convert the observations into profiles of cloud fraction and cloud water content. We will also describe a new algorithm for deriving liquid water path from radiometers which avoids the problems of producing lwp in the absence of water clouds and negative lwp values. These profiles can then be averaged into hourly values with a vertical resolution corresponding to the grid of the NWP model such as ECMWF, Met Office, MeteoFrance, KNMI RACMO and SMHI RCA. The model and observed values can be viewed at http://www.met.reading.ac.uk/radar/cloudnet/quicklooks/. Monthly and annual statistics of the performance of the NWP models in representing cloud fraction, ice and liquid water content are then produced over several European observatories and the ARM sites.
This poster will be displayed at the ARM Science Team Meeting.