A102-04
Likelihood of Cloud Gaps Allowing C-gas Exchange Estimates: A Survey of Amazonia Across Broad Temporal Scales, Cloud-Gap Dimensions, and Accuracy Assessments

Thursday, 10 December 2020: 16:28
Virtual
Robert B Chatfield, NASA Ames Research Center, Moffett Field, CA, United States, Robert F Esswein, Bay Area Environmental Research Institute Moffett Field, NASA Ames Earth Science, Moffett Field, CA, United States, Alexei Lyapustin, NASA Goddard Space Flight Center, Greenbelt, MD, United States, Yujie Wang, University of Maryland Baltimore County, Baltimore, MD, United States, Steven T Massie, Natl Ctr Atmospheric Research, Boulder, CO, United States, Christopher O'Dell, Colorado State University, Fort Collins, CO, United States, Sean Crowell, University of Oklahoma, Norman, OK, United States and The GeoCarb Science Team
Recorded Presentation
Abstract:
We present a first analysis of cloud gaps in the Amazon aspiring to increase the likelihood that a variably situated sensor footprint will provide an acceptably accurate estimate of XCH4 and XCO2using current or forthcoming systems (Sentinel5p “Tropomi”, OCO-3, GeoCarb, Sentinel 5, and several modest and small-sized satellites). Acceptable samples from such cloudy areas are scarce but are vital to the understanding of methane and carbon dioxide fluxes since bottom-up simulations of the region exhibit significant differences.

The analysis summarizes products of the multi-angle MAIAC algorithm using the long record of the NASA MODIS satellite sensors and a sample of the detailed hourly description available from the NOAA GOES-16 advanced baseline imager. Cloud gaps of sufficient size are necessary since light path measured using the O2 A-band at 0.76 µm is used as a guide to the light path at 1.61 µm, 2.06 µm, and 2.32 µm. For the fraction-of-a-percent accuracies required to understand fluxes, relatively small effects (aerosol scattering and multiple reflections of light rays off of cloud sides and the surface) do affect retrieved light-path between 0.76 µm and the other wavelengths. It is extraordinarily cost effective to understand error budgets for all satellites and to plan targeting strategies especially for scene-selecting satellites.

The meteorology of cloud cover is intricate; consequently we will describethe cloud gaps with multiple categories: longitude, latitude, hour-of-day, multi-day periods, seasonal, and interannual variations. Additionally, it is important to address the trade-off between estimated accuracy of individual retrievals as a function of gap size: we explore the decrease of opportunities when we require gap sizes described by lat-lon variations of rectangles dimensioned 8x8 km, 16x16 km, and 24x24 km. At the larger dimensions, the footprint of the sensor matters surprisingly little between 0.1 km-capable sensors and 6-km-capable sensors: the distance to the nearest cloud becomes dominant. We progress to correlate the role of albedo and aerosol optical depth with the cloud-gap parameters. Finally, we briefly summarize others’ current understanding of the effects of cloud-distance on accuracy and systematic methods to improve the accuracy in presence of clouds.

<< Previous Abstract | Next Abstract >>