Explore Research Products in the PAR
It may take a few hours for recently added research products to appear in PAR search results.


Title: Exploring the Use of Machine Learning to Improve Vertical Profiles of Temperature and Moisture
Abstract Vertical profiles of temperature and dewpoint are useful in predicting deep convection that leads to severe weather, which threatens property and lives. Currently, forecasters rely on observations from radiosonde launches and numerical weather prediction (NWP) models. Radiosonde observations are, however, temporally and spatially sparse, and NWP models contain inherent errors that influence short-term predictions of high impact events. This work explores using machine learning (ML) to postprocess NWP model forecasts, combining them with satellite data to improve vertical profiles of temperature and dewpoint. We focus on different ML architectures, loss functions, and input features to optimize predictions. Because we are predicting vertical profiles at 256 levels in the atmosphere, this work provides a unique perspective at using ML for 1D tasks. Compared to baseline profiles from the Rapid Refresh (RAP), ML predictions offer the largest improvement for dewpoint, particularly in the middle and upper atmosphere. Temperature improvements are modest, but CAPE values are improved by up to 40%. Feature importance analyses indicate that the ML models are primarily improving incoming RAP biases. While additional model and satellite data offer some improvement to the predictions, architecture choice is more important than feature selection in fine-tuning the results. Our proposed deep residual U-Net performs the best by leveraging spatial context from the input RAP profiles; however, the results are remarkably robust across model architecture. Further, uncertainty estimates for every level are well calibrated and can provide useful information to forecasters.  more » « less
Award ID(s):
2019758
PAR ID:
10512785
Author(s) / Creator(s):
; ; ; ;
Publisher / Repository:
AMS Journals
Date Published:
Journal Name:
Artificial Intelligence for the Earth Systems
Volume:
3
Issue:
1
ISSN:
2769-7525
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; (, EGUsphere)
    Abstract. Global Navigation Satellite Systems (GNSS) radio occultation (RO) is one of the most vital remote sensing techniques globally and of major importance for numerical weather prediction (NWP) and climate science. However, retrieving profiles of atmospheric quantities such as temperature or humidity from GNSS observations is not straightforward and dedicated algorithms still have their limitations. One of these limitations is the need for external meteorological data in the retrieval process. Various new RO missions have led to an enormous increase in data amounts and with over 10000 globally-distributed, daily profiles, RO can be considered big data nowadays. In this study, we make use of this fact by developing a new retrieval method based on a deep learning model, which only needs RO-specific quantities as an input to produce atmospheric profiles. The model is trained on almost a full year of data from COSMIC-2 and Spire RO missions, using vertical profiles of bending angle (BA) and other RO parameters as input features and operational results from a standard retrieval algorithm as target values for supervised learning. Initial results from both internal and external validation using reanalysis and radiosonde data suggest that this method produces results with an accuracy comparable to standard algorithms, while mitigating the need for external information in the retrieval process itself. These initial results serve as a starting point for further development of data-driven models for RO, which could significantly enhance the quality of RO products utilized in, e.g., climate sciences by mitigating external biases and increasing independence from other techniques. 
    more » « less
  2. ; ; ; ; (, Remote Sensing)
    This paper describes, along with some validation results, the one-dimensional variational method (1D-Var) that is in use at the University Corporation for Atmospheric Research (UCAR) to retrieve atmospheric profiles of temperature, pressure, and humidity from the observation of the Global Navigation Satellite System (GNSS) radio occultation (RO). The retrieved profiles are physically consistent among the variables and statistically optimal as regards to a priori error statistics. Tests with idealized data demonstrate that the 1D-Var is highly effective in spreading the observational information and confirm that the method works as designed and expected, provided that correct input data are given. Tests for real-world data sets show that the retrieved profiles agree remarkably well with global weather analyses and collocated high vertical resolution radiosonde observations, and that the 1D-Var can produce value-added retrievals with respect to a priori profiles. We also find that the retrieved profiles are of exceptional long-term stability, suggesting that the 1D-Var can provide an excellent climate data record. 
    more » « less
  3. ; ; ; ; ; ; ; ; (, Atmospheric Measurement Techniques)
    Abstract. Remote sensing measurements have been widely used to estimate the planetary boundary layer height (PBLHT). Each remote sensing approach offers unique strengths and faces different limitations. In this study, we use machine learning (ML) methods to produce a best-estimate PBLHT (PBLHT-BE-ML) by integrating four PBLHT estimates derived from remote sensing measurements at the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Southern Great Plains (SGP) observatory. Three ML models – random forest (RF) classifier, RF regressor, and light gradient-boosting machine (LightGBM) – were trained on a dataset from 2017 to 2023 that included radiosonde, various remote sensing PBLHT estimates, and atmospheric meteorological conditions. Evaluations indicated that PBLHT-BE-ML from all three models improved alignment with the PBLHT derived from radiosonde data (PBLHT-SONDE), with LightGBM demonstrating the highest accuracy under both stable and unstable boundary layer conditions. Feature analysis revealed that the most influential input features at the SGP site were the PBLHT estimates derived from (a) potential temperature profiles retrieved using Raman lidar (RL) and atmospheric emitted radiance interferometer (AERI) measurements (PBLHT-THERMO), (b) vertical velocity variance profiles from Doppler lidar (PBLHT-DL), and (c) aerosol backscatter profiles from micropulse lidar (PBLHT-MPL). The trained models were then used to predict PBLHT-BE-ML at a temporal resolution of 10 min, effectively capturing the diurnal evolution of PBLHT and its significant seasonal variations, with the largest diurnal variation observed over summer at the SGP site. We applied these trained models to data from the ARM Eastern Pacific Cloud Aerosol Precipitation Experiment (EPCAPE) field campaign (EPC), where the PBLHT-BE-ML, particularly with the LightGBM model, demonstrated improved accuracy against PBLHT-SONDE. Analyses of model performance at both the SGP and EPC sites suggest that expanding the training dataset to include various surface types, such as ocean and ice-covered areas, could further enhance ML model performance for PBLHT estimation across varied geographic regions. 
    more » « less
  4. ; ; ; ; ; ; ; ; (, npj Clean Air)
    Abstract As wildfires increase in frequency and intensity, accurately representing the vertical distribution of smoke in numerical models is critical for assessing impacts to air quality, but remains highly uncertain. In this study, we leverage satellite retrievals of total column carbon monoxide (CO) and aerosol layer height (ALH) to evaluate two state-of-the-art regionals and global models, one using a plume rise parameterization to estimate smoke injection height (RAP-Chem) and another placing smoke at the surface (MOMO-Chem). We introduce a novel metric that utilizes the differing vertical sensitivities of two satellite sensors observing CO (TROPOMI and CrIS) to infer the vertical distribution of wildfire smoke using a joint CO column ratio. We find that RAP-Chem better captures the distribution of CO and ALH related to the 2020 western US megafire event than MOMO-Chem. However, RAP-Chem underestimates surface CO concentrations, revealing that current plume rise parameterizations are limited in their ability to partition smoke correctly in the vertical column. These results show that synergistic use of satellite data can provide additional constraints on the vertical distribution of smoke, thus providing insights into the strengths and limitations of current plume rise parameterizations and a pathway to improvement. 
    more » « less
  5. ; ; ; ; ; ; ; (, Atmospheric Measurement Techniques)
    Abstract. One-dimensional variational retrievals of temperature and moisture fields from hyperspectral infrared (IR) satellite sounders use cloud-cleared radiances (CCRs) as their observation. These derived observations allow the use of clear-sky-only radiative transfer in the inversion for geophysical variables but at reduced spatial resolution compared to the native sounder observations. Cloud clearing can introduce various errors, although scenes with large errors can be identified and ignored. Information content studies show that, when using multilayer cloud liquid and ice profiles in infrared hyperspectral radiative transfer codes, there are typically only 2–4 degrees of freedom (DOFs) of cloud signal. This implies a simplified cloud representation is sufficient for some applications which need accurate radiative transfer. Here we describe a single-footprint retrieval approach for clear and cloudy conditions, which uses the thermodynamic and cloud fields from numerical weather prediction (NWP) models as a first guess, together with a simple cloud-representation model coupled to a fast scattering radiative transfer algorithm (RTA). The NWP model thermodynamic and cloud profiles are first co-located to the observations, after which the N-level cloud profiles are converted to two slab clouds (TwoSlab; typically one for ice and one for water clouds). From these, one run of our fast cloud-representation model allows an improvement of the a priori cloud state by comparing the observed and model-simulated radiances in the thermal window channels. The retrieval yield is over 90%, while the degrees of freedom correlate with the observed window channel brightness temperature (BT) which itself depends on the cloud optical depth. The cloud-representation and scattering package is benchmarked against radiances computed using a maximum random overlap (RMO) cloud scheme. All-sky infrared radiances measured by NASA's Atmospheric Infrared Sounder (AIRS) and NWP thermodynamic and cloud profiles from the European Centre for Medium-Range Weather Forecasts (ECMWF) forecast model are used in this paper. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Text Encoded Conversion
  • XML
  • Save / Share this Record
  • Send to Email