Articles | Volume 14, issue 3
https://doi.org/10.5194/amt-14-1941-2021
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
https://doi.org/10.5194/amt-14-1941-2021
© Author(s) 2021. This work is distributed under
the Creative Commons Attribution 4.0 License.
the Creative Commons Attribution 4.0 License.
Using machine learning to model uncertainty for water vapor atmospheric motion vectors
Joaquim V. Teixeira
CORRESPONDING AUTHOR
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA
Hai Nguyen
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA
Derek J. Posselt
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA
Hui Su
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA
Longtao Wu
Jet Propulsion Laboratory, California Institute of Technology, Pasadena, USA
Related authors
No articles found.
Juan Socuellamos, Matthew Lebsock, Raquel Rodriguez Monje, Marcin Kurowski, Derek Posselt, and Robert Beauchamp
EGUsphere, https://doi.org/10.5194/egusphere-2025-4248, https://doi.org/10.5194/egusphere-2025-4248, 2025
This preprint is open for discussion and under review for Atmospheric Measurement Techniques (AMT).
Short summary
Short summary
Using simulated airborne and spaceborne W-band (94 GHz) and G-band (239 GHz) radar observations, this article presents an optimal estimation framework that includes dual-frequency reflectivity and differential absorption measurements to obtain more accurate drizzle retrievals. We demonstrate that an improvement of more than one order of magnitude in the retrieval of the drizzle mixing ratio, droplet concentration, and mass-weighted diameter can be achieved compared to W-band only results.
Hai Nguyen, Derek Posselt, Igor Yanovsky, Longtao Wu, and Svetla Hristova-Veleva
Atmos. Meas. Tech., 17, 3103–3119, https://doi.org/10.5194/amt-17-3103-2024, https://doi.org/10.5194/amt-17-3103-2024, 2024
Short summary
Short summary
Accurate global wind estimation is crucial for weather prediction and environmental modeling. Our study investigates a method to refine atmospheric motion vectors (AMVs) by comparing them with high-precision active-sensor winds. Leveraging supervised learning, we discovered that using high-precision active-sensor data can significantly reduce biases in passive-sensor winds in addition to providing estimates of the wind errors, thereby improving their reliability.
Jonathan Hobbs, Matthias Katzfuss, Hai Nguyen, Vineet Yadav, and Junjie Liu
Geosci. Model Dev., 17, 1133–1151, https://doi.org/10.5194/gmd-17-1133-2024, https://doi.org/10.5194/gmd-17-1133-2024, 2024
Short summary
Short summary
The cycling of carbon among the land, oceans, and atmosphere is a closely monitored process in the global climate system. These exchanges between the atmosphere and the surface can be quantified using a combination of atmospheric carbon dioxide observations and computer models. This study presents a statistical method for investigating the similarities and differences in the estimated surface–atmosphere carbon exchange when different computer model assumptions are invoked.
Vijay Natraj, Ming Luo, Jean-Francois Blavier, Vivienne H. Payne, Derek J. Posselt, Stanley P. Sander, Zhao-Cheng Zeng, Jessica L. Neu, Denis Tremblay, Longtao Wu, Jacola A. Roman, Yen-Hung Wu, and Leonard I. Dorsky
Atmos. Meas. Tech., 15, 1251–1267, https://doi.org/10.5194/amt-15-1251-2022, https://doi.org/10.5194/amt-15-1251-2022, 2022
Short summary
Short summary
High-fidelity monitoring and forecast of air quality and the hydrological cycle require understanding the vertical distribution of temperature, humidity, and trace gases at high spatiotemporal resolution. We describe a new instrument concept, called the JPL GEO-IR Sounder, that would provide this information for the first time from a single instrument platform. Simulations demonstrate the benefits of combining measurements from multiple wavelengths for this purpose from geostationary orbit.
Sudip Chakraborty, Jonathon H. Jiang, Hui Su, and Rong Fu
Atmos. Chem. Phys., 21, 12855–12866, https://doi.org/10.5194/acp-21-12855-2021, https://doi.org/10.5194/acp-21-12855-2021, 2021
Short summary
Short summary
Boreal autumn is the main wet season over the Congo basin. Thus, changes in its onset date have a significant impact on the rainforest. This study provides compelling evidence that the cooling effect of aerosols modifies the timing and strength of the southern African easterly jet that is central to the boreal autumn wet season over the Congo rainforest. A higher boreal summer aerosol concentration is positively correlated with the boreal autumn wet season onset timing.
Cited articles
Bormann, N., Hernandez-Carrascal, A., Borde, R., Lutz, H. J., Otkin, J. A., and Wanzong, S.: Atmospheric motion vectors from model simulations. Part I:
Methods and characterization as single-level estimates of wind, J.
Appl. Meteorol. Clim., 53, 47–64, https://doi.org/10.1175/JAMC-D-12-0336.1, 2014.
Breiman, L.: Random forests, Mach. Learn., 45, 5–32, 2001.
Cassola, F. and Burlando, M.: Wind speed and wind energy forecast through
Kalman filtering of Numerical Weather Prediction model output, Appl.
Energ., 99, 154–166, 2012.
Coulston, J. W., Blinn, C. E., Thomas, V. A., and Wynne, R. H.:
Approximating prediction uncertainty for random forest regression
models, Photogramm. Eng. Rem. S., 82, 189–197, 2016.
Fraley, C. and Raftery, A. E.: MCLUST: Software for model-based clustering,
density estimation and discriminant analysis (No. TR-415), Washington
University, Seattle Department of Statistics, Seattle, Washington, USA, 2002.
Fraley, C., Raftery, A. E., Murphy, T. B., and Scrucca, L: mclust version 4 for R: normal mixture modeling for model-based clustering, classification, and density estimation, Washington University, Seattle Department of Statistics, Seattle, Washington, USA, 2012.
Gneiting, T. and Katzfuss, M.: Probabilistic forecasting, Annu. Rev. Stat. Appl., 1, 125–151, 2014.
Hernandez-Carrascal, A. and Bormann, N.: Atmospheric motion vectors from
model simulations. Part II: Interpretation as spatial and vertical averages
of wind and role of clouds, J. Appl. Meteorol. Clim.,
53, 65–82, 2014.
Holmlund, K.: The utilization of statistical properties of satellite-derived atmospheric motion vectors to derive quality indicators, Weather Forecast., 13, 1093–1104, https://doi.org/10.1175/1520-0434(1998)013<1093:TUOSPO>2.0.CO;2, 1998.
Holmlund, K., Velden, C. S., and Rohn, M.: Enhanced automated quality
control applied to high-density satellite-derived winds, Mon. Weather
Rev., 129, 517–529, 2001.
Kawa, S. R., Erickson, D. J., Pawson, S., and Zhu, Z.: Global CO2 transport simulations using meteorological data from the NASA data assimilation system, J. Geophys. Res.-Atmos., 109, D18312,
https://doi.org/10.1029/2004JD004554, 2004.
Kwon, Y., Won, J. H., Kim, B. J., and Paik, M. C.: Uncertainty
quantification using Bayesian neural networks in classification: Application
to biomedical image segmentation, Comput. Stat. Data An., 142, 106816, https://doi.org/10.1016/j.csda.2019.106816, 2020.
Le Marshall, J., Rea, A., Leslie, L., Seecamp, R., and Dunn, M.: Error
characterisation of atmospheric motion vectors, Aust. Meteorol.
Mag., 53, 123–131, 2004.
Mueller, K. J., Wu, D. L., Horváth, Á., Jovanovic, V. M., Muller, J. P., Di Girolamo, L., Garay, M. J., Diner, D. J., Moroney, C. M., and Wanzong, S.: Assessment of MISR cloud motion vectors (CMVs) relative to GOES and MODIS
atmospheric motion vectors (AMVs), J. Appl. Meteorol.
Clim., 56, 555–572, https://doi.org/10.1175/JAMC-D-16-0112.1, 2017.
Nguyen, H., Cressie, N., and Hobbs, J.: Sensitivity of Optimal
Estimation Satellite Retrievals to Misspecification of the Prior Mean and
Covariance, with Application to OCO-2 Retrievals, Remote Sensing, 11, 2770, https://doi.org/10.3390/rs11232770, 2019.
Posselt, D. J., Wu, L., Mueller, K., Huang, L., Irion, F. W., Brown, S., Su, H., Santek, D., and Velden, C. S.: Quantitative Assessment of State-Dependent
Atmospheric Motion Vector Uncertainties, J. Appl. Meteorol. Clim., 58, 2479–2495, https://doi.org/10.1175/JAMC-D-19-0166.1, 2019.
Putman, W., da Silva, A. M., Ott, L. E., and Darmenov, A.: Model Configuration for the 7-km GEOS-5 Nature Run, Ganymed Release (Non-hydrostatic 7 km Global Mesoscale Simulation), GMAO Office Note No. 5 (Version 1.0), Global Modeling and Assimilation Office, NASA Goddard Space Flight Center, Greenbelt, Maryland, USA, 18, available at: https://opendap.nccs.nasa.gov/dods/OSSE/G5NR/Ganymed/7km/0.0625_deg (last access: 30 September 2019), 2014.
Salonen, K., Cotton, J., Bormann, N., and Forsythe, M.: Characterizing AMV
Height-Assignment Error by Comparing Best-Fit Pressure Statistics from the
Met Office and ECMWF Data Assimilation Systems, J. Appl. Meteorol. Clim.,
54, 225–242, https://doi.org/10.1175/JAMC-D-14-0025.1, 2015.
Staffell, I. and Pfenninger, S.: Using bias-corrected reanalysis to simulate
current and future wind power output, Energy, 114, 1224–1239, 2016.
Swail, V. R. and Cox, A. T.: On the use of NCEP–NCAR reanalysis surface
marine wind fields for a long-term North Atlantic wave hindcast, J.
Atmos. Ocean. Tech., 17, 532–545, 2000.
Tibshirani, R. J. and Efron, B.: An Introduction to the
Bootstrap, Mg. Stat. Pro., 57, 1–436, 1993.
Tran, D., Dusenberry, M., van der Wilk, M., and Hafner, D.: Bayesian
layers: A module for neural network uncertainty, in: Advances in Neural Information Processing Systems, Conference on Neural Information Processing Systems, 8–14 December 2019, Vancouver, Canada, 14660–14672, 2019.
Tripathy, R. K. and Bilionis, I.: Deep UQ: Learning deep neural network
surrogate models for high dimensional uncertainty quantification, J. Comput. Phys., 375, 565–588, 2018.
Velden, C. S. and Bedka, K. M.: Identifying the Uncertainty in Determining
Satellite-Derived Atmospheric Motion Vector Height Attribution, J. Appl.
Meteorol. Clim., 48, 450–463, https://doi.org/10.1175/2008JAMC1957.1, 2009.
Zeng, X., Ackerman, S., Ferraro, R. D., Lee, T. J., Murray, J. J., Pawson, S., Reynolds, C., and Teixeira, J.: Challenges and opportunities in NASA weather research, B. Am. Meteorol. Soc., 97, 137–140, 2016.
Short summary
Wind-tracking algorithms produce atmospheric motion vectors (AMVs) by tracking satellite observations. Accurately characterizing the uncertainties in AMVs is essential in assimilating them into data assimilation models. We develop a machine-learning-based approach for error characterization which involves Gaussian mixture model clustering and random forest using a simulation dataset of water vapor, AMVs, and true winds. We show that our method improves on existing AMV error characterizations.
Wind-tracking algorithms produce atmospheric motion vectors (AMVs) by tracking satellite...