of 13
GEOPHYSICAL RESEARCH LETTERS
Supporting Information for ”Data Drought in the
Humid Tropics: How to Overcome the Cloud Barrier
in Greenhouse Gas Remote Sensing”
C. Frankenberg
1
,
2
,
, Y. M. Bar-On
1
,
, Y. Yin
3
, P.O. Wennberg
1
,
4
, D.J.
Jacob
5
, A.M. Michalak
6
,
7
1
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California, USA
2
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, California, USA
3
Department of Environmental Studies, New York University, New York, New York, USA
4
Division of Engineering and Applied Science, California Institute of Technology, Pasadena, California, USA
5
School of Engineering and Applied Science, Harvard University, Cambridge, Massachusetts, USA
6
Carnegie Institution for Science, Stanford, California, USA
7
Department of Earth System Science, Stanford University, Stanford, California, USA
Authors contributed equally to the study
Contents of this file
1. Text S1 to S4
2. Figures S1 to S8
Introduction
March 26, 2024, 8:26pm
X - 2
:
Our overall analysis was performed in the major tropical areas but many of the figures
in the main text only showed the Amazon. Here we provide a broader scope with a focus
on Africa and Southeast Asia as well. We also investigate potential optima in terms of
footprint sizes. We will show a visual guide of our analysis, a brief investigation into the
time-of-day dependency using OCO-3, as well as an evaluation of optimal footprint sizes.
Text S1: Visual representation.
Figure S1 provides a visual overview of our data analysis. Sentinel-2 images are acquired
every 5 days, resulting in 6 snapshots per month. These are displayed as matrix in the left
part of the figure, showing a 10x10km
2
area for clarity. For our analysis, we use the entire
image size within 1x1 degrees. The left column shows Sentinel 2 reflectances as false-color
image, with the reflectance at 835, 1610, and 2200 nm, respectively in red, green and blue.
These bands are close to the OCO-2 bands, thus the red-ness of the scene indicated much
higher O
2
A-band albedo compared to the CO
2
bands.
More than half of the scenes show typical shallow-cumulus cloud fields, which would
greatly benefit from smaller ground-pixels for GHG measurements. Only five are mostly
cloud-free, even for larger ground pixels than OCO-2 provides. A 2x2km area is indicated
in blue in the first column. The visualization supports the skewed distribution observed in
our statistical analysis. This results in long data-free periods during the wet season. The
right column shows our corresponding binary cloud filter, i.e. black areas are cloud-free
while white ones are either clouds or cloud shadows. Only one out of 72 images covering
a 10x10km
2
area is entirely cloud free.
Text S2: Impact of time-of-day on clouds and data yields.
March 26, 2024, 8:26pm