A Machine Learning Approach to the Observation Operator for Satellite Radiance Data Assimilation

Liang, Jianyu; Terasaki, Koji; Miyoshi, Takemasa

doi:10.2151/jmsj.2023-005

Cited by 9 publications

(4 citation statements)

References 33 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…They were also used to emulate 3D effects on RT (Meyer et al, 2022) and to generate near-infrared satellite images (ρ 1 : 1.6 µm) in (Baur et al, 2023). Recently, Liang et al (2023) used ML to assimilate different bands of BT from Advanced Microwave Sounding Unit-A. In their framework, the satellite observed radiance was assimilated using RTTOV and specific MLP models were trained for each band and satellite.…”

Section: Radiative Transfer Modelsmentioning

confidence: 99%

Focal-TSMP: Deep learning for vegetation health prediction and agricultural drought assessment from a regional climate simulation

Shams Eddin,

Gall

2023

Preprint

View full text Add to dashboard Cite

Abstract. In this study, we investigate applying deep learning (DL) models on a regional climate simulation produced by the Terrestrial Systems Modelling Platform (TSMP Ground to Atmosphere G2A) for vegetation health modeling and agricultural drought assessment. The TSMP simulation is performed in a free mode and the DL model is used in an intermediate step to synthesize Normalized Difference Vegetation Index (NDVI) and Brightness Temperature (BT) images from the TSMP simulation over Europe. These predicted images are then used to derive different vegetation and drought indices like NDVI anomaly, BT anomaly, Vegetation Condition Index (VCI), Thermal Condition Index (TCI), and Vegetation Health Index (VHI). To ensure reliability and to assess the model applicability with different seasonality and spatial variability, we provide an analysis of model biases and uncertainties across different regions over the Pan-Europe domain. We further provide an analysis about the contribution of the input variables from the TSMP model components to ensure a better understanding of the model prediction. A comprehensive evaluation on the long-term TSMP using reference remote sensing data showed sufficiently good agreements between the model predictions and observations. While model performance varies on the test set between different climate regions, it achieves a mean absolute error (MAE) of 0.027 and 1.90 K° with coefficient of determination (R2) scores of 0.88 and 0.92 for NDVI and BT, respectively, at 0.11° resolution for sub-seasonal predictions. Our study could be used as a complimentary evaluation framework for climate change simulations with TSMP. Moreover, the developed DL model could be integrated with data assimilation and used for down-stream tasks, i.e., modelling the impact of extreme events on vegetation responses with different climate change scenarios. In summary, we demonstrate the feasibility of using DL on a TSMP simulation to synthesize NDVI and BT, which can be used for agricultural drought forecasting. Our implementation is publicly available at the project page (https://hakamshams.github.io/Focal-TSMP).

show abstract

Section: Radiative Transfer Modelsmentioning

confidence: 99%

Focal-TSMP: Deep learning for vegetation health prediction and agricultural drought assessment from a regional climate simulation

Shams Eddin,

Gall

2023

Preprint

View full text Add to dashboard Cite

show abstract

“…The observation operator, which can be another computational bottleneck, has also been targeted using ML (Geer, 2021; Jung et al., 2010; X. Liang et al., 2022; J. Liang et al., 2023; Stegmann et al., 2022; Wang et al., 2022). Other efforts have used ML methods to identify regions of tropical cyclone activity to target high‐resolution modeling in a subdomain (Lee et al., 2019), to perform bias correction of model forecasts before they are fed into the assimilation algorithm (Arcucci et al., 2021; Chen et al., 2022), and apply time‐varying localization to the covariance structure of the system (Lacerda et al., 2021).…”

Section: Introductionmentioning

confidence: 99%

“…The observation operator, which can be another computational bottleneck, has also been targeted using ML (Geer, 2021;Jung et al, 2010;X. Liang et al, 2022;J. Liang et al, 2023;Stegmann et al, 2022;Wang et al, 2022).…”

mentioning

confidence: 99%

A Machine Learning Augmented Data Assimilation Method for High‐Resolution Observations

Howard,

Subramanian,

Hoteit

2024

J Adv Model Earth Syst

View full text Add to dashboard Cite

The accuracy of initial conditions is an important driver of the forecast skill of numerical weather prediction models. Increases in the quantity of available measurements, particularly high‐resolution remote sensing observational data products from satellites, are valuable inputs for improving those initial condition estimates. However, the traditional data assimilation methods for integrating observations into forecast models are computationally expensive. This makes incorporating dense observations into operational forecast systems challenging, and it is often prohibitively time‐consuming. Additionally, high‐resolution observations often have correlated observation errors which are difficult to estimate and create problems for assimilation systems. As a result, large quantities of data are discarded and not used for state initialization. Using the Lorenz‐96 system for testing, we demonstrate that a simple machine learning method can be trained to assimilate high‐resolution data. Using it to do so improves both initial conditions and forecast accuracy. Compared to using the Ensemble Kalman Filter with high‐resolution observations ignored, our augmented method has an average root‐mean‐squared error reduced by 37%. Ensemble forecasts using initial conditions generated by the augmented method are more accurate and reliable at up to 10 days of forecast lead time.

show abstract

“…The observation operator, which can be another computational bottleneck, has also been targeted using ML (Jung et al, 2010;J. Liang et al, 2023;Wang et al, 2022;Geer, 2021;X.…”

Section: Introductionmentioning

confidence: 99%

A Machine Learning Augmented Data Assimilation Method for High-Resolution Observation

Howard

Subramanian

Hoteit

2023

Preprint

View full text Add to dashboard Cite

The accuracy of initial conditions is an important driver of the forecast skill of numerical weather prediction models. Increases in the quantity of available measurements, in particular high-resolution remote sensing observational data products from satellites, are valuable inputs for improving those initial condition estimates. However, the data assimilation methods used for integrating observations into forecast models are computationally expensive. This makes incorporating dense observations into operational forecast systems challenging, and it is often prohibitively time-consuming. As a result, large quantities of data are discarded and not used for state initialization. We demonstrate, using the Lorenz-96 system for testing, that a simple machine learning method can be trained to assimilate high-resolution data. Using it to do so improves both initial conditions and forecast accuracy. Compared to using the Ensemble Kalman Filter with high-resolution observations ignored, our augmented method has an average root-mean-squared error reduced by 15%. Ensemble forecasts using initial conditions generated by the augmented method are more accurate and reliable at up to 10 days of forecast lead time.

show abstract

A Machine Learning Approach to the Observation Operator for Satellite Radiance Data Assimilation

Abstract: This pre-publication manuscript may be downloaded, distributed and used under the provisions of the Creative Commons Attribution 4.0 International (CC BY 4.0) license.

Cited by 9 publications

References 33 publications

Focal-TSMP: Deep learning for vegetation health prediction and agricultural drought assessment from a regional climate simulation

Focal-TSMP: Deep learning for vegetation health prediction and agricultural drought assessment from a regional climate simulation

A Machine Learning Augmented Data Assimilation Method for High‐Resolution Observations

A Machine Learning Augmented Data Assimilation Method for High-Resolution Observation

Contact Info

Product

Resources

About