Hyperspectral Cloud Top Retrievals

Robert E. Holz, Steve Ackerman, Matthew McGill,

Paolo Antonelli and Fred Nagle

Overview

Part 1. CO2 Sorting + Slicing cloud height retrieval validation

Part 2. Night time polar cloud height detection using AIRS hyperspectral measurements

CO2 Channel Selection Algorithm (CO2 Sorting)

The Sorted Clear Sky Spectrum

High and Thick Cloud

CO2 Sorting: Sensitivity to Brightness Temperature

High and Thick Cloud

Thinner Cloud

High and Thick Cloud

Thinner Cloud

Low Cloud

Selected CO2 ChannelsHigh Cloud

Mid Level Cloud

February 22, 2003

S-HIS - CPL February 22, 2003

S-HIS - CPL Cloud OD < 1.0

(km)

S-HIS Cloud Top OD Sensitivity

.

Altitu

de

0.2 0.4

0.6

0.8 1.

0 1.2

CPL OD Contours

SHIS - CPL

SHIS Cloud Top Retrieval

S-HIS Cloud Top OD

Distribution of S-HIS Cloud Level Optical Depth February 22

CPL optical depth at the S-HIS cloud height

Part 2: Night Time Polar Cloud

Height Retrievals Using AIRS

Hyperspectal Measurements

One BT could be 3 different cloud heights

Polar Cloud Height Retrieval Problem

AIRS Polar Cloud Height Retrieval 14:00 UTC

AIRS Polar Cloud Height Retrieval 22:00 UTC

Conclusions and Future Plans

• The combined CO2 slicing + sorting cloud top retrieval improves the CO2 slicing results for optically thin clouds

• Lidar retrieved integrated cloud optical depth is a more representative measure of cloud top retrieval performance.

• It is possible to detect temperature inversions above arctic stratus using hyperspectral measurements

• The inversion information reduces the uncertainty in the arctic stratus IR cloud top retrievals

• Future:

– Apply the CO2 sorting + slicing retrieval to AIRS satellite data

– Integrate the CO2 sorting + slicing retrieval with the arctic inversion detection