Cloud liquid and ice water content estimation from satellite: a regression approach based on neural networks

Cloud microphysics in terms of their liquid/ice water content and particle size are the principal factors addressed to study and understand the behavior behind the climate change phenomenon. Based on remotely sensed measurements, in the last decades, some evidence exists that an increase in temperature leads to an increase in cloud liquid water content (CLWC). The temperature dependence of ice water content (CIWC) is also evident from measurements of midlatitude cirrus clouds. Hence, innovative methods, such as those based on the use of Artificial Intelligence (AI) allowing a more relevant investigation of how clouds influence the hydrological cycle and radiative components of the Earth's climate system, are required. This work investigates the capability of a statistical regression scheme of CLWC and CIWC, implemented through the use of a multilayer feed-forward neural network (NN). The whole methodology is applied to a set of simulated IASI-NG L1C and MWS acquisitions, covering the global scale. The NN regression analysis shows good agreement with the test data. The retrieved cloud liquid water and ice profiles have an accuracy of 20 to 60% depending on the given layer. Finally, the layer with the maximum concentration is accurately identified.

[1]  Jorge Cadima,et al.  Principal component analysis: a review and recent developments , 2016, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[2]  William L. Smith,et al.  Retrieval of atmospheric profiles and cloud properties from IASI spectra using super-channels , 2009 .

[3]  Christopher D. Barnet,et al.  Hyperspectral Earth Observation from IASI: Five Years of Accomplishments , 2012 .

[4]  Rolf Müller,et al.  From ERA-Interim to ERA5: considerable impact of ECMWF's next-generation reanalysis on Lagrangian transport simulations , 2018, Atmospheric Chemistry and Physics.

[5]  Peter I. Frazier,et al.  A Tutorial on Bayesian Optimization , 2018, ArXiv.

[6]  Philip W. Rosenkranz,et al.  Cloud liquid‐water profile retrieval algorithm and validation , 2006 .

[7]  Sida I. Wang,et al.  Dropout Training as Adaptive Regularization , 2013, NIPS.

[8]  Jimmy Ba,et al.  Adam: A Method for Stochastic Optimization , 2014, ICLR.

[9]  W. Paul Menzel,et al.  Retrieval of Cloud Parameters from Satellite Sounder Data: A Simulation Study , 1989 .

[10]  Patrick Minnis,et al.  Estimation of water cloud properties from satellite microwave , 1998 .

[11]  Klaus-Robert Müller,et al.  Efficient BackProp , 2012, Neural Networks: Tricks of the Trade.

[12]  Carlo Luschi,et al.  Revisiting Small Batch Training for Deep Neural Networks , 2018, ArXiv.

[13]  Guigang Zhang,et al.  Deep Learning , 2016, Int. J. Semantic Comput..

[14]  Vincent Guidard,et al.  Towards IASI-New Generation (IASI-NG): impact of improved spectral resolution and radiometric noise on the retrieval of thermodynamic, chemistry and climate variables , 2013 .

[15]  Guido Masiello,et al.  EOF regression analytical model with applications to the retrieval of atmospheric temperature and gas constituents concentration from high spectral resolution infrared observations , 2013 .

[16]  Jasper Snoek,et al.  Practical Bayesian Optimization of Machine Learning Algorithms , 2012, NIPS.

[17]  F. Rabier,et al.  Toward the improvement of short‐range forecasts by the analysis of cloud variables from IASI radiances , 2014 .

[18]  Nitish Srivastava,et al.  Dropout: a simple way to prevent neural networks from overfitting , 2014, J. Mach. Learn. Res..

[19]  Lutz Prechelt,et al.  Early Stopping - But When? , 2012, Neural Networks: Tricks of the Trade.

[20]  Nitish Srivastava,et al.  Improving Neural Networks with Dropout , 2013 .

[21]  C. Bohren,et al.  An introduction to atmospheric radiation , 1981 .

[22]  Pietro Mastro,et al.  Characterization of the Observational Covariance Matrix of Hyper-Spectral Infrared Satellite Sensors Directly from Measured Earth Views , 2020, Sensors.

[23]  N. Bormann,et al.  An update on the RTTOV fast radiative transfer model (currently at version 12) , 2018 .

[24]  Fuzhong Weng,et al.  One-dimensional variational retrieval algorithm of temperature, water vapor, and cloud water profiles from advanced microwave sounding unit (AMSU) , 2005, IEEE Transactions on Geoscience and Remote Sensing.

[25]  Frank S. Marzano,et al.  Flower Constellation of Millimeter-Wave Radiometers for Tropospheric Monitoring at Pseudogeostationary Scale , 2009, IEEE Transactions on Geoscience and Remote Sensing.