GRWG/GDWG Web Meeting 2023-07-06

GSICS IR Sub-Group Web Meeting

Agenda

1. Tim Hewison (EUMETSAT): Investigation of Inter-calibration Bias over Hot Land (20 minutes + 10 minutes discussion)

2. Wenhui Wang (UMD): Inter-calibration of VIIRS and CrIS (20 minutes)

3. Likun Wang (UMD): Collocating VIIRS with CrIS using only VIIRS Terrain Corrected Geolocation Datasets (20 minutes)

Attendees

Chair: Likun Wang

NOAA: Likun Wang, Fred Wu, Wehhui Wang, Bomin Huang, Denis Tremblay, Manik Bali

EUMETSAT: Tim Hewison, Vincent Debaecker, Ali Mousivand, Mounir Lekouara, Viju John

CMA: Chengli Qi, Xingwei He, ZW Wang

University of Wisconsin: Dave Tobin

KMA: Euidong Hwang

JMA: Kazuki Kodera, Masaya Takahashi, Misaki Eiki

Univ. of Hamburg: Martin Burgdorf

ISRO: Pradeep Thapliyal

ESA: Stefano Casadio

NASA/GFSC: Tiejun Chang

University of Reading: Jon Mittaz

Summary

Tim Hewison (EUMETSAT): Investigation of Inter-calibration Bias over Hot Land (20 minutes + 10 minutes discussion)

Tim Hewison, Vincent Debaecker, and Ali Mousivand from EUMETSAT reported recent progress on the investigation of the warm bias over hot land from GEO-LEO and LEO-LEO inter-calibration. It was caused by the asymmetric time sample difference when collocating observations from two satellites. They highlighted the importance of symmetric time sampling for collocated observations, especially over land. They further checked the similar issues for LEO and LEO inter-calibration, including SLSTR-IASI and AVHRR-IASI. The following discussion suggested that all the agencies check the time different pattern for the collocation data.

Manik Bali: Are SLSTA-IASI inter-calibration nadir or off-nadir views?

R: It includes high zenith angle data.

Likun Wang: Would you like to explain how the asymmetric time sample difference happened when collocating observations from two satellites?

R: This is caused by a code bug in the algorithms, and we fixed that. We encourage other agencies to check time difference collocation patterns.

Dave Tobin (Comments): In our recent paper , we created symmetric time distribution by filtering out outlier data.

Fred Wu (comments): We use GEO-LEO collocation datasets for navigation check. For example, using VIIRS check geolocation of GOES. We are concerned that cloud moving (from time difference of collocations) can cause the bias. We use the temporal interpolation.

Fred Wu: If the temperature change rate scale determines the bias, did you check the CrIS inter-calibration data, which is on 1330 orbit and has small temperature change rate.

R: We have not check the CrIS data. In answer to Fred's earlier question - Vincent just checked the sensitivity of SEVIRI-CrIS relative biases to time differences in the daytime over land - it is almost flat (slope -0.042K/min for IR10.8).

Sensitivity of SEVIRI-CrIS to dt in daytime over land is channel-dependent:

-0.042K/min for IR10.8

-0.007K/min in IR12.0

-0.055K/min for IR13.4

Masaya (comments): I checked our GEO-LEO inter-calibration data and found that time difference is symmetrically distributed. But I will further investigate it.

CMA and KMA will investigate it.

Wenhui Wang (UMD): Inter-calibration of VIIRS and CrIS (20 minutes)

Wenhui Wang reported the inter-calibration results between CrIS and VIIRS from a different perspective, which treats CrIS as a reference and then evaluates VIIRS calibration quality. She also applied the CrIS gap filling methods and found that the inter-calibration between CrIS and VIIRS is greatly improved for VIIRS M13 and M14 bands.

Likun Wang (comments): On Slide 6, there is a jump at M15 bands when CrIS was switched from normal spectral resolution to full resolution. It should not change too much because the radiometric calibration and spectral coverage does not change and only spectral resolution change.
Dave Tobin (comments): I will also to check it.

Dave (comments): CrIS /NOAA-21 is still under calibration improvements. The updated polarization improvements will effect on the M13 for scene and angle bias. That will be implemented in September 2023. In addition, I will share CrIS -VIIRS inter-calibration results with you.

Fred Wu: Slide 8, why is there orbit variation?
R: It is caused by scene temperature dependent.

Likun Wang (UMD): Collocating VIIRS with CrIS using only VIIRS Terrain Corrected Geolocation Datasets (20 minutes)

VIIRS currently releases both Terrain Corrected (TC) and Non-TC geolocation datasets. Under the plan, the Non-TC geolocation datasets will be turned off in the future. However, other sensors on the same platform(e.g. CrIS) only have Non-TC geolocation datasets. Likun Wang recently investigated how this impacts on the collocation and proposed a resolution to mitigate the effects.

Cheng-Li Qi: Geolocation datasets from the sounding instrument (like ATMS and CrIS) are terrain corrected or not?

R: They are not.

Tim: How is this going to affect on the GEO-LEO inter-calibration?

R: Currently most geolocation datasets on geostationary are still Non-TC. So far, it has no impacts.

Fred Wu: How to collocate two sensors based on the view angles?

Dave (comments): The first peer-reviewed publication on spatial collocation is the work by Nagle and Holz (2009), which provides a guidance for a general methodology that can be applied to a wide range of satellite, aircraft, and surface measurements and allow for efficient collocation with measurements having varying spatial and temporal sampling.

Nagle, F.W.; Holz, R.E. Computationally efficient methods of collocating satellite, aircraft, and ground observations. J. Atmos. Oceanic Technol. 2009, 26, 1585–1595.

Outcome

Action: ??