Browsing by Author "Chen, Haonan, advisor"

Now showing 1 - 5 of 5

Open Access
Deep learning for radar beam blockage correction
(Colorado State University. Libraries, 2023) Tan, Songjian, author; Chen, Haonan, advisor; Chandrasekaran, V., committee member; Wang, Haonan, committee member
This thesis aims to propose a deep learning framework based on generative adversarial networks (GANs) for correcting partial beam blockage regions in polarimetric radar observations. The correction of such data is an essential step in radar data quality control and subsequent quantitative applications, especially in complex terrain environments. The proposed methodology is demonstrated using two S-band operational Weather Surveillance Radar - 1988 Doppler (WSR-88D) located in different regions of the western United States, characterized by different precipitation types. To train the GAN model, observation sectors of both radars are manually cropped to simulate partial beam blockage situations. The effectiveness of the trained models is demonstrated using independent precipitation events in Texas and California, and their generalization capacity is examined by cross-testing the data with different precipitation features. The beam blockage correction performance is compared with a traditional linear interpolation approach, and the results show that the proposed approach significantly improves the continuity of precipitation observations in both domains. While visible discrepancies exist between the models trained based on convective and stratiform precipitation events in Texas and California, respectively, both models outperform the traditional interpolation method. The repaired observations demonstrate great potential for improved quantitative applications, despite the unavailability of ground truth for real blocked radar data.
Open Access
Deep learning for short-term prediction of wildfire using geostationary satellite observations
(Colorado State University. Libraries, 2024) Saqer, Yousef, author; Chen, Haonan, advisor; Azimi-Sadjadi, Mahmood R., committee member; Wei, Yu, committee member
The aim of this thesis is to utilize the Geostationary Operational Environmental Satellite (GOES) data for predictions regarding the intensity and potential path of wildfires. Using GOES to identify wildfires and extracting data from those events to help train a deep learning model. Three fires were selected for training the deep learning model: the Sequoia, Calwood, and Maui fires. The GOES data of the fires was obtained from band 7 which operates in the Shortwave Window or 3.9μm wavelength, band 7 is able to capture hotspots which is beneficial for wildfire prediction. The radiance data from band 7 is pulled from an Amazon Web Service (AWS) and becomes part of a dataset of 2513 samples. The data is then stacked to form a time series of approximately two hours and converted into a compressed h5 file. The pipeline distributes the dataset by taking in twenty five minutes of input data and feeding four different models to predict seventy five minutes, one hundred minutes, and one hundred and twenty five minutes of data. The data is then fed into a deep learning model utilizing a model known as Self Attention Gated Recurrent Unit (SaGRU). The SaGRU is tested four times, once for predicting seventy five minutes, once for predicting one hundred minutes, and twice for one hundred and twenty five minutes. The models were then compared against each other regarding Mean Squared Error (MSE) and Mean Absolute Error (MAE) along with the Normalized Mean Squared Error (NME) and the Normalized Mean Absolute Error (NMAE). Each metric was taken along multiple thresholds comparing the performance when hotspots are present and when hotspots are absent. The resultant showed that regardless of the sequence length, there was minimal negative impact on early predictions, but as the predicted sequence increased significant loss could be seen on the later predicted frames.
Open Access
High-rate GNSS satellite clock estimation: implications for radio occultation bending angle precision
(Colorado State University. Libraries, 2024) Ko, Yao-Chun, author; Chen, Haonan, advisor; Yao, Jian, advisor; Chiu, Christine, committee member
The Global Navigation Satellite System (GNSS) radio occultation (RO) technique plays a vital role in collecting data for meteorological and space weather prediction. It is exemplified by the COSMIC-2 low-Earth-orbit (LEO) satellite constellation, which collects the GNSS signals from an elevation angle of 90° to below the horizon. Those GNSS observation data above 5° elevation angle are used for the precise orbit determination of satellites, while those GNSS observation data below 5° are used for the RO processing. A key part of the RO processing is to estimate the bending angle due to the atmospheric refraction, which requires an accurate information of the positions and clock offsets of both the transmitter (i.e., GNSS satellite) and the receiver (i.e., COSMIC-2 satellite). Previous research at University Corporation for Atmospheric Research (UCAR) [1] indicates a notable reduction in the intrinsic uncertainty of GLONASS radio occultation when employing higher-rate GNSS satellite clock products (e.g., from 30-second sampling interval to 2-second sampling interval). However, that work only analyzed one day of dataset. To analyze multiple days of dataset, I have developed a software program that can automatically generate high-rate GNSS clock products by using a GNSS toolkit called GINAN [2]. This program is also important to the future UCAR's RO postprocessing and near-real-time processing. To be specific, it first downloads, merges, and decimates 1-second GNSS-receiver data from 50 worldwide ground stations, and then runs the GINAN software to generate clock products. I have validated the clock products generated by the program by comparing to International GNSS Service (IGS) analysis centers' clock products – the standard deviation of the time difference between our clock products and the clock products published by the Center for Orbit Determination in Europe is as small as ~ 0.1 nanoseconds. Using one week of 2-sec clock products generated by the program, I have run the standard RO processing and found that the bending-angle uncertainty of the GLONASS RO has been reduced by ~ 34%, as compared to if using the existing 30-sec clock products. Admittedly, there is no obvious improvement for the GPS RO because the GPS satellite clocks are stable at a short term of <= 30 seconds. By pushing down the noise of the RO technique, we can possibly observe the atmosphere at an unprecedented precision which could benefit the research of atmosphere modelling, the operation of weather monitoring and forecast, and even the study of space weather.
Open Access
Improving radar quantitative precipitation estimation through optimizing radar scan strategy and deep learning
(Colorado State University. Libraries, 2024) Wang, Liangwei, author; Chen, Haonan, advisor; Chandrasekaran, Venkatchalam, committee member; Wang, Haonan, committee member
As radar technology plays a crucial role in various applications, including weather forecasting and military surveillance, understanding the impact of different radar scan elevation angles is paramount to optimize radar performance and enhance its effectiveness. The elevation angle, which refers to the vertical angle at which the radar beam is directed, significantly influences the radar's ability to detect, track, and identify targets. The effect of different elevation angles on radar performance depends on factors such as radar type, operating environment, and target characteristics. To illustrate the impact of lowering the minimum scan elevation angle on surface rainfall mapping, this article focuses on the KMUX WSR-88D radar in Northern California as an example, within the context of the National Weather Service's efforts to upgrade its operational Weather Surveillance Radar. By establishing polarimetric radar rainfall relations using local disdrometer data, the study aims to estimate surface rainfall from radar observations, with a specific emphasis on shallow orographic precipitation. The findings indicate that a lower scan elevation angle yields superior performance, with a significant 16.1% improvement in the normalized standard error and a 19.5% enhancement in the Pearson correlation coefficient, particularly for long distances from the radar. In addition, conventional approaches to radar rainfall estimation have limitations, recent studies have demonstrated that deep learning techniques can mitigate parameterization errors and enhance precipitation estimation accuracy. However, training a model that can be applied to a broad domain poses a challenge. To address this, the study leverages crowdsourced data from NOAA and SFL, employing a convolutional neural network with a residual block to transfer knowledge learned from one location to other domains characterized by different precipitation properties. The experimental results showcase the efficacy of this approach, highlighting its superiority over conventional fixed-parameter rainfall algorithms. Machine learning methods have shown promising potential in improving the accuracy of quantitative precipitation estimation (QPE), which is critical in hydrology and meteorology. While significant progress has been made in applying machine learning to QPE, there is still ample room for further research and development. Future endeavors in machine learning-based QPE will primarily focus on enhancing model accuracy, reliability, and interpretability while considering practical operational applications in hydrology and meteorology.
Embargo
Interpolating RGB radar images based on machine learning
(Colorado State University. Libraries, 2023) Yi, Chenke, author; Chandrasekar, V., advisor; Chen, Haonan, advisor; Siller, Thomas, committee member; Gooch, Steven, committee member
Weather radar interpolation is the process of estimating and predicting rainfall data in areas that are not directly observed by radar. This technique is commonly used in weather forecasting, flood prediction, and agricultural planning. The main goal of weather radar interpolation is to produce accurate and reliable precipitation maps in areas with limited radar coverage or where the radar data is incomplete. The interpolation methods can be categorized into two main groups: deterministic and stochastic. Deterministic methods use mathematical equations and physical models to estimate the rainfall, while stochastic methods rely on statistical algorithms to analyze the correlations between the radar measurements and ground observations. In recent years, machine learning algorithms have also been applied to weather radar interpolation, showing promising results in accuracy and robustness. In this paper, we mainly propose a radar image interpolation method based on spatio-temporal convolutional networks. The experiments are mainly compared and analyzed for different combinations of networks, connection methods, and different loss functions.