Steve Ackerman, R. Holz, R Frey, S. Platnick, A. Heidinger, and a bunch of others.

Outline Using CALIOP to Validate MODIS Cloud Detection, Cloud Height Assignment, Optical Properties Clouds and Surface Low level (<5 Km cloud tops) Marine Clouds

Horizontal Sampling of MODIS

Vertical sampling CALIOP

Validation of cloud and clear sky detection AUGUST 2006

Zonal Means July 2007 June 2008 High agreement for ocean scenes Cold land night a problem 6

Agreement 2007-2012 100% 85% Hit Rate (%): ( (#agree clear pixels + #agree cloudy pixels) / total pixels) * 100.0 1-km MODIS cloud mask is compared to 1-km CALIOP data for JJA 2007-2012. University of Wisconsin - Madison is 2 degrees Map resolution

Validation, validation, validation One month, regional and Skill scores

Hit rate and false alarm CLOUD

Hit rate and false alarm CLEAR

Cloud fraction as a function of scene Cloud amounts at 1 km scale from CALIOP and MODIS Water/Land over South America

Cloud Top Altitude

black CALIPSO, red MODIS CO2 August 2006 The mean cloud top height differences (MODIS CALIOP).The bottom image presents the standard deviation of the MODIS CALIOP cloud height differences for each 5-degree region.

Derived cloud top altitude comparison 0.05 Normalized Number of Occurrence 0.045 All Cloudy FOV CALIPSO < 5 km CALIPSO > 5 km 0.04 0.035 0.03 0.025 0.02 0.015 0.01 0.005 0 20 15 10 5 0 Cloud Height Difference (MODIS CALIPSO (km)) 5 As expected, for thin clouds, the MODIS (IR passive approach) is sensitive to a layer below the physical cloud top.

CALIPSO/MODIS Comparison black CALIPSO, red MODIS CO2 Comparison is helping to understand performance of MODIS CO2 Cooperative Institute for Meteorological Satellite Studies Slicing derived cloud altitudes. Black points are MODIS cloud heights.

Cloud Optical Depth

Cirrus Optical Depth Single Layer Ice Clouds MODIS NVIS IOT 6.0e+00 3.5 5.0e+00 3 4.1e+00 3.3e+00 2.5 2.7e+00 2 2.2e+00 1.5 1.8e+00 1 1.5e+00 0.5 1.2e+00 1 0.5 1 1.5 2 2.5 CALIOP OD 3 3.5

Cirrus Optical Depth: MODIS C5, CALIOP and IR Single Layer Ice Clouds Single Layer Ice Clouds 1.5e+02 4.5 4.1e+00 3.5 3.3e+00 3 2.7e+00 2.5 2.2e+00 2 1.8e+00 1.5 1.5e+00 1 1.2e+00 0.5 1 0.5 1 1.5 2 2.5 3 3.5 4 4.5 LBLDIS IR Extinction IOT CALIOP Non-Constrained IOT 4 MODIS NVIS IOT 4.5 5.0e+00 9.0e+01 4 5.5e+01 3.5 3.3e+01 3 2.0e+01 2.5 1.2e+01 2 7.4e+00 1.5 4.5e+00 1 2.7e+00 0.5 1.6e+00 1 1 2 3 LBLDIS IR Extinction IOT 4

Cirrus Optical Depth Single Layer Ice Clouds 5.0e+00 3.5 4.1e+00 MODIS NVIS IOT 3 3.3e+00 2.5 2.7e+00 2 2.2e+00 1.5 1.8e+00 1 1.5e+00 0.5 1.2e+00 1 0.5 1 1.5 2 2.5 LBLDIS IR Extinction OD 3 3.5

IR Cloud Optical Depth Comparison

Cirrus Optical Depth Single Layer Ice Clouds 0.15 MODIS CALIOP Unconstrained CALIOP Constrained 3.5 MODIS NVIS IOT 3 2.5 4.5e+00 2 2.7e+00 1.5 1 Normalized Frequency 7.4e+00 0.1 0.05 1.6e+00 0.5 1 0.5 1 1.5 2 2.5 3 3.5 CALIOP OD 0-40 -30-20 Improvements made for Collection 6-10 0 10 20 Calculated - Measured 11um (K) 30 40

Clouds In the Boundary Layer

Boundary Layer Heights JJA CALIPSO PBL HEIGHT 3 km 2 km 1 km

Boundary Layer Height DJF JJA

Cloud Top variations Day, Dry Season Croplands South America Day, Wet Season Croplands South America A histogram analysis of CALIOP derived cloud top altitude in km above the surface, for cloud tops less than 8 km over South America for ecosystems classified as Croplands for the dry season (left) and the wet season (right) during daytime.

Cloud microphysics

MODIS, CALIOP and CloudSat together CALIOP Cloud Phase CALIOP Backscatter CALIOP Depolarization CloudSat radar refelctivity CloudSat precipitation flag

Occurrence of CALIOP oriented Ice Crystals One year, ocean areas and cloud top below 5 km

Seasonal Distribution HOIC One year, ocean areas and cloud top below 5 km N.H. Orientated S.H. Orientated

HOIC and CloudSat Precipitation Rain Snow Mixed Atlantic Rain Snow Mixed Pacific

What have we learned, and are learning! The average fractional agreement for MODIS IFOVs with CALIOP is on the order of 90%. " Reference to other cloud detection methods) " Passive approach problematic for polar night.! MODIS CO2 cloud height biases of cirrus are lower than CALIOP!!!! altitudes by ~ 2.5 km, consistent with expectations for comparison with lidar. MODIS cloud height biases for marine stratus. MODIS and CALIOP optical depths are consistent with each other and IR approaches. Cloud properties coupled to surface types. A unique 1-year data set for horizontally oriented ice crystals

Thank you its fun working with these data.