Observed and modelled long-term water cloud statistics for the Murg Valley Kerstin Ebell, Susanne Crewell, Ulrich Löhnert Institute for Geophysics and.

Observed and modelled long-term water cloud statistics for the Murg Valley

Kerstin Ebell, Susanne Crewell, Ulrich Löhnert

Institute for Geophysics and Meteorology, University of Cologne

7th COPS Workshop, 27 – 29 October 2008, Strasbourg

2 7th COPS Workshop, 27 – 29 October 2008, Strasbourg 27th COPS Workshop, 27 – 29 October 2008, Strasbourg

Motivation 30 July 2007, Murg Valley

IR sky temperature + LWP

N W S E N

UTC

14

13

12

11

• LWC is important

- as prognostic parameter in NWP/climate models- in describing cloud radiative effects- for predicting attenuation in radio propagation problems

due to scarsity of observations simple cloud models were developed to diagnose LWC profiles from radiosonde measurements

• Clouds are complex in - time/space- microphysics

• Observations need to captureall characteristics see Poster by Kneifel et al. for spatio-temporal structure

• Liquid water content (LWC)can not be observed by a single instrument Integrated Profiling Technique

3 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Motivation

How are water clouds characterized in the Murg valley during April – December 2007?

„observations“

How are they represented by models?

Comparison to• simple cloud models• (NWP model COSMO-DE)

Cloudnet Target Classification(provided by Ewan O‘Connor)

Integrated ProfilingTechnique

4 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

provides continuous vertical profiles of temperature, humidity and liquid water content (LWC); first attempts to derive droplet number concentration

data set of retrieved profiles available for April – December 2007!

Integrated Profiling Technique (IPT) – Overview

Microwave radiometers Cloud radar A priori information

(radiosondes)

CloudNet Target Categorization: Each pixel is categorized in terms of

the presence of liquid droplets, ice, insects or aerosol.

Discrimination between precipitating and non-precipitating clouds.

Integration of measurements in the framework of optimal estimation:

inversion method including - instrument characteristics (theory, error) - forward model, i.e. radiative transfer, Z- LWC-relationship - background information

Retrieval

5 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Cloud models - overview

Simple cloud models: diagnosis of clouds from radiosonde profiles Modified adiabatic approach (Karstens et al., 1994) Decker model (Decker et al., 1978) modified Salonen model (Mattioli et al., 2006)

General principle:

1) Localization of cloud base and cloud top height

2) Calculation of LWC: depending on height over cloud base (Mod.ad., Salonen) depending on cloud thickness (Decker), Min 0.05 gm-3, Max 0.2 gm-3

Assumption: no water clouds below -20 °C

RH sondeRhcrit modad / DeckerRhcrit Salonen 06

6 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Results – Critical humidity thresholds

Distribution of relative humidity (radiosonde) in cloudypixels (April – December 2007);

cloudy pixels: Cloudnet classification detects cloud droplets

detection of clouds from relative humidity profiles problematic

7 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Results – Cloudnet and cloud models

Number of clouds

Cloud occurence (%)

Mod. Ad. 69.4

Decker 69.4

Salonen 75.3

Cloudnet 55.5

April-December 2007

Cloudnet classification profiles within 30 min. aftera RS ascent are taken into account for the comparison

8 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Results – Cloud base height (lowest cloud)

April-December 2007

9 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Results – Cloud thickness distribution

April-December 2007

most frequent cloud thickness is 100-200 m

only ~17 % of clouds with thicknesses > 500 m

only ~10 % of clouds with thicknesses > 1000 m

10 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Results - LWP

Cloud freq. (%)

Mean LWP (gm-2)

Mod. Ad. 56.3 41.2

Decker 56.3 36.3

Salonen 65.8 14.6

IPT 34.1 15.9

COSMO-DE

21.4 20.8

For COSMO-DE evaluation more time steps and inclusion of fractional clouds needed!

11 7th COPS Workshop, 27 – 29 October 2008, Strasbourg

Results - LWC

12 7th COPS Workshop, 27 – 29 October 2008, Strasbourg 127th COPS Workshop, 27 – 29 October 2008, Strasbourg

Summary

• Derivation of liquid water, temperature and humidity profiles for nearly the whole AMF deployment (21.4.-31.12.2007) using the Integrated Profiling Technique 88110 profiles (cloudy: 33168) corresponding to 2706 hours (5 min sampling)

• Most clouds are single layer, 100-200 m thick with low LWP

• Simple cloud models based on RS profiles - overestimate the occurence of water clouds by ~25-35%- overestimate thin clouds (< 300 m)- overestimate of multilayer clouds and LWP

• COSMO-Model evaluation is limited due to missing information on fractional cloudiness

Outlook• Investigation of cloud radiative effects using IPT/Cloudnet data and single

column version of the radiation code of the COSMO model

• sensitivity studies, e.g. additional retrieval of number concentration