Using Observations from NASA’s CAMP2Ex Campaign to Evaluate Smoke Optical Properties in GEOS
In the Fall of 2019, NASA’s P3 Orion aircraft participated in 19 research flights in the Philippines region aimed at observing the interaction of aerosols, clouds, and radiation within the atmosphere as part of the Cloud, Aerosol and Monsoon Processes Philippines Experiment (CAMP2Ex) field campaign. During the first half of the campaign, the large-scale circulation was characterized by southwesterly monsoonal flow that transported smoke from fires in Indonesia across the Sulu Sea and towards the Philippines, coupled by convection typical of a tropical environment. Much can be learned from the wealth of observation collected in the complex regime, but the data provides a valuable resource for evaluating aerosols in models such as the Goddard Earth Observing System (GEOS). Within GEOS, the sources, sinks, and chemistry of aerosols are handled within the Goddard Chemistry Aerosol Radiation and Transport (GOCART) model, which recently underwent a series of science and code upgrades. As part of the upgrades, a distinction was made between organic aerosol emitted through biomass burning and anthropogenic sources as the two types of organic aerosol can differ in their optical properties. Here, optical properties of biomass burning aerosol are assessed using a transect from a CAMP2Ex flight through a smoke plume.
The NASA Langley Aerosol Research Group Experiment (LARGE) suite of instruments collected in situ observations of aerosol mass, size distribution, and optical properties. Fine mode aerosols were passed into chambers that dried and rehydrated the aerosol to determine how much radiation would be scattered under varying levels of relative humidity. As humidity increases, hydrophilic aerosols, including those present within smoke, swell, scattering more radiation than if they were dry. Figure 1 shows scatterplots comparing scattering and extinction at visible wavelengths between GEOS and the LARGE observations, colored based on either the bias in mass concentration for organic carbon or relative humidity. As show in panels a and b, GEOS suffers from a positive bias in the mass of organic carbon. Much of this bias is introduced through the data assimilation process aimed at correcting the total column extinction of aerosols, or aerosol optical depth, according to satellite observations from MODIS. For both the mass concentration and aerosol optical depth to be accurate, GEOS will need to properly represent the optical properties of each aerosol species.
Under dry conditions, GEOS underestimates both scattering and extinction when there is not an overabundance of organic carbon (Figure 1a and b). Conversely, under ambient conditions, GEOS overestimates extinction, with excessive variability in the range of extinction (Figure 1c). Some of the overestimation in ambient extinction is associated with a bias in relative humidity. Using the observed relative humidity, the optics calculations were performed again to eliminate the influence of moisture as shown in panel d. While some of the bias is reduce, GEOS still overestimates extinction at 532 nm when the relative humidity is observation corrected.
Figure 1: Scatter plot of observations from LARGE versus GEOS GOCART2G for (a) dry scattering at 550 nm, (b) dry extinction at 532 nm, (c) ambient extinction at 532 nm, and (d) ambient extinction at 532 nm with bias corrected relative humidity in GEOS. Panels (a) and (b) are colored based on the bias in organic carbon mass concentration and panels (c) and (d) are colored based on the bias in relative humidity.
Although ambient relative humidity dictates the optical properties used to establish the column aerosol optical depth, LARGE measures scattering at a relative humidity around 40% and 80% and infers a relationship between relative humidity and the amount of aerosol scattering to arrive at the ambient scattering. Figure 2 shows scattering from the LARGE observations and GEOS, corrected for the relative humidity bias, as a function of relative humidity. Below a relative humidity of 80%, GEOS performs well in terms of the mean and variability for scattering at 550 nm. However, GEOS assumes an exponential relationship between scattering and relative humidity with the observations indicating a linear relationship, resulting in an overestimation in scattering that increases with relative humidity. Given the extrapolation in the observations when the relative humidity exceeds 80%, uncertainty remains as to whether there are biases present in the observations, model, or both. This advocates for additional techniques for deriving the relationship between aerosol scattering and relative humidity, such as using a high spectral resolution lidar (HSRL).
Figure 2: Scatter plot of aerosol scattering at 550 nm from CAMP2Ex observations and the GEOS model as a function of observed relative humidity.
