Extracting Taklimakan Dust Parameters from AIRS with Artificial Neural Network Method

Abstract

Two back-propagation artificial neural network retrieval models have been developed for obtaining the dust aerosol optical depth (AOD) and dust-top height (DTH), respectively, from Atmospheric InfraRed Sounder (AIRS) brightness temperature (BT) measurements over Taklimakan Desert area. China Aerosol Remote Sensing Network (CARSNET) measurements at Tazhong station were used for dust AOD validation. Results show that the correlation coefficient of dust AODs between AIRS and CARSNET reaches 0.88 with a deviation of −0.21, which is the same correlation coefficient as the AIRS dust AOD and the Moderate-Resolution Imaging Spectroradiometer (MODIS) aerosol optical depth (AOD) product. In the AIRS DTH retrieval model, there is an option to include the collocated MODIS deep blue (DB) AOD as additional input for daytime retrieval; the independent dust heights from Cloud-Aerosol Lidar with Orthogonal Polarization (CALIOP) are used for AIRS DTH validation, and results show that the DTHs derived from the combined AIRS BT measurements and MODIS DB AOD product have better accuracy than those from AIRS BT measurements alone. The correlation coefficient of DTHs between AIRS and independent CALIOP dust heights is 0.79 with a standard deviation of 0.41 km when MODIS DB AOD product is included in the retrieval model. A series of case studies from different seasons were examined to demonstrate the feasibility of retrieving dust parameters from AIRS and potential applications. The method and approaches can be applied to process measurements from advanced infrared (IR) sounder and high-resolution imager onboard the same platform.

RETRACTED: Seasonal correlation of aerosols with soil moisture, evapotranspiration, and vegetation over Pakistan using remote sensing

“…The diverse change in global climate patterns due to cooling and warming of the atmosphere is because of aerosols present in the atmosphere. However, having an understanding of aerosols and their interaction has become tough due to the variations in their distribution within the atmosphere [ 14 ]. Lensky and Rosenfeld [ 15 ] studied the effect of convective clouds and rainfall on aerosols.…”

Dust Radiative Effect Characteristics during a Typical Springtime Dust Storm with Persistent Floating Dust in the Tarim Basin, Northwest China

“…The MODIS instruments onboard the NASA Terra and Aqua satellites work with more accurate spatial-temporal resolutions with a 0.1 × 0.1 • horizontal resolution in this study and supply almost daily the whole global coverage and monitoring of the Earth's changes, including changes in tropospheric aerosols [44,45]. Ascribed to the difficulty in retrieving aerosol characteristics on the desert, as well as clouds, MODIS satellite observations corresponding to the simulation period had a restricted spatial-temporal coverage area [46]. The dark-blue algorithm in MODIS is more appropriate for desert areas, allowing for the identification of dust storms and the accurate retrieval of the characteristics of dust aerosols over the desert [47,48].…”

Four‐Dimensional Wind Fields From Geostationary Hyperspectral Infrared Sounder Radiance Measurements With High Temporal Resolution

“…While previous methodologies on wind retrievals are mostly based on objective detection and tracking of movements, the new ML based algorithm developed in this study is training a feed‐forward neuron network (NN) with numerous channels of BT as inputs and the matched‐up wind profiles from ERA5 reanalysis as true values. NNs have shown significant ability to fit intricate nonlinear relationships between inputs and outputs through training and optimizations (Feng et al., 2017; LeCun et al., 2015; Van Gerven & Bohte, 2017), and have been successful in a number of meteorological applications with satellite observations (Boukabara et al., 2019; Milstein & Blackwell, 2016; Tao et al., 2018; Whitburn et al., 2016; Yao et al., 2019; Zhou & Grassotti, 2020). A feed‐forward NN with an input layer, an output layer, and two hidden layers in between is used to retrieve the 3D U/V‐wind profiles from hyperspectral observations from GIIRS.…”