dor_id: 4134772

506.#.#.a: Público

590.#.#.d: Los artículos enviados a la revista "Atmósfera", se juzgan por medio de un proceso de revisión por pares

510.0.#.a: Consejo Nacional de Ciencia y Tecnología (CONACyT); Sistema Regional de Información en Línea para Revistas Científicas de América Latina, el Caribe, España y Portugal (Latindex); Scientific Electronic Library Online (SciELO); SCOPUS, Web Of Science (WoS); SCImago Journal Rank (SJR)

561.#.#.u: https://www.atmosfera.unam.mx/

561.#.#.a: no

650.#.4.x: Físico Matemáticas y Ciencias de la Tierra

336.#.#.b: article

336.#.#.3: Artículo de Investigación

336.#.#.a: Artículo

351.#.#.6: https://www.revistascca.unam.mx/atm/index.php/atm/index

351.#.#.b: Atmósfera

351.#.#.a: Artículos

harvesting_group: RevistasUNAM

270.1.#.p: Revistas UNAM. Dirección General de Publicaciones y Fomento Editorial, UNAM en revistas@unam.mx

590.#.#.c: Open Journal Systems (OJS)

270.#.#.d: MX

270.1.#.d: México

590.#.#.b: Concentrador

883.#.#.u: https://revistas.unam.mx/catalogo/

883.#.#.a: Revistas UNAM

590.#.#.a: Coordinación de Difusión Cultural

883.#.#.1: https://www.publicaciones.unam.mx/

883.#.#.q: Dirección General de Publicaciones y Fomento Editorial

850.#.#.a: Universidad Nacional Autónoma de México

856.4.0.u: https://www.revistascca.unam.mx/atm/index.php/atm/article/view/53053/46921

100.1.#.a: Wen, Wu; Li, Lei; Chan, P. W.; Liu, Yuan-yuan; Wei, Min

524.#.#.a: Wen, Wu, et al. (2023). Research on the usability of different machine learning methods in visibility forecasting. Atmósfera; Vol. 37, 2023; 99-111. Recuperado de https://repositorio.unam.mx/contenidos/4134772

245.1.0.a: Research on the usability of different machine learning methods in visibility forecasting

502.#.#.c: Universidad Nacional Autónoma de México

561.1.#.a: Instituto de Ciencias de la Atmósfera y Cambio Climático, UNAM

264.#.0.c: 2023

264.#.1.c: 2023-01-17

653.#.#.a: visibility forecast; deep learning; machine learning; time-series forecasting

506.1.#.a: La titularidad de los derechos patrimoniales de esta obra pertenece a las instituciones editoras. Su uso se rige por una licencia Creative Commons BY-NC 4.0 Internacional, https://creativecommons.org/licenses/by-nc/4.0/legalcode.es, para un uso diferente consultar al responsable jurídico del repositorio por medio del correo electrónico editora@atmosfera.unam.mx

884.#.#.k: https://www.revistascca.unam.mx/atm/index.php/atm/article/view/53053

001.#.#.#: 022.oai:ojs.pkp.sfu.ca:article/53053

041.#.7.h: eng

520.3.#.a: Haze pollution, mainly characterized by low visibility, is one of the main environmental problems currently faced by China. Accurate haze forecasts facilitate the implementation of preventive measures to control the emission of air pollutants and thereby mitigate haze pollution. However, it is not easy to accurately predict low visibility events induced by haze, which requires not only accurate prediction for weather elements, but also refined and real-time updated source emission inventory. In order to obtain reliable forecasting tools, this paper studies the usability of several popular machine learning methods, such as support vector machine (SVM), k-nearest neighbor, and random forest, as well as several deep learning methods, on visibility forecasting. Starting from the main factors related to visibility, the relationships between wind speed, wind direction, temperature, humidity, and visibility are discussed. Training and forecasting were performed using the machine learning methods. The accuracy of these methods in visibility forecasting was confirmed through several parameters (i.e., root-mean-square error, mean absolute error, and mean absolute percentage error). The results show that: (1) among all meteorological parameters, wind speed was the best at reflecting the visibility change patterns; (2) long short-term memory recurrent neural networks (LSTM RNN), and gated recurrent unit (GRU) methods perform almost equally well on short-term visibility forecasts (i.e., 1, 3, and 6 h); (3) a classical machine learning method (i.e., the SVM) performs well in mid- and long-term visibility forecasts; (4) machine learning methods also have a certain degree of forecast accuracy even for long time periods (e.g., 7 2h).

773.1.#.t: Atmósfera; Vol. 37 (2023); 99-111

773.1.#.o: https://www.revistascca.unam.mx/atm/index.php/atm/index

022.#.#.a: ISSN electrónico: 2395-8812; ISSN impreso: 0187-6236

310.#.#.a: Trimestral

300.#.#.a: Páginas: 99-111

264.#.1.b: Instituto de Ciencias de la Atmósfera y Cambio Climático, UNAM

doi: https://doi.org/10.20937/ATM.53053

handle: 578d2c03b0647c4c

harvesting_date: 2023-06-20 16:00:00.0

856.#.0.q: application/pdf

file_creation_date: 2023-01-19 15:09:05.0

file_modification_date: 2023-01-19 15:09:07.0

file_name: fd0a7b8074ba98662bd1746ecf7b842d1a07d3540347042242a3779e5389e915.pdf

file_pages_number: 13

file_format_version: application/pdf; version=1.4

file_size: 2770045

last_modified: 2023-06-20 16:00:00

license_url: https://creativecommons.org/licenses/by-nc/4.0/legalcode.es

license_type: by-nc

No entro en nada

No entro en nada 2

Artículo

Research on the usability of different machine learning methods in visibility forecasting

Wen, Wu; Li, Lei; Chan, P. W.; Liu, Yuan-yuan; Wei, Min

Instituto de Ciencias de la Atmósfera y Cambio Climático, UNAM, publicado en Atmósfera, y cosechado de Revistas UNAM

Licencia de uso

Procedencia del contenido

Entidad o dependencia
Instituto de Ciencias de la Atmósfera y Cambio Climático, UNAM
Revista
Atmósfera
Repositorio
Revistas UNAM
Contacto
Revistas UNAM. Dirección General de Publicaciones y Fomento Editorial, UNAM en revistas@unam.mx

Cita

Wen, Wu, et al. (2023). Research on the usability of different machine learning methods in visibility forecasting. Atmósfera; Vol. 37, 2023; 99-111. Recuperado de https://repositorio.unam.mx/contenidos/4134772

Descripción del recurso

Autor(es)
Wen, Wu; Li, Lei; Chan, P. W.; Liu, Yuan-yuan; Wei, Min
Tipo
Artículo de Investigación
Área del conocimiento
Físico Matemáticas y Ciencias de la Tierra
Título
Research on the usability of different machine learning methods in visibility forecasting
Fecha
2023-01-17
Resumen
Haze pollution, mainly characterized by low visibility, is one of the main environmental problems currently faced by China. Accurate haze forecasts facilitate the implementation of preventive measures to control the emission of air pollutants and thereby mitigate haze pollution. However, it is not easy to accurately predict low visibility events induced by haze, which requires not only accurate prediction for weather elements, but also refined and real-time updated source emission inventory. In order to obtain reliable forecasting tools, this paper studies the usability of several popular machine learning methods, such as support vector machine (SVM), k-nearest neighbor, and random forest, as well as several deep learning methods, on visibility forecasting. Starting from the main factors related to visibility, the relationships between wind speed, wind direction, temperature, humidity, and visibility are discussed. Training and forecasting were performed using the machine learning methods. The accuracy of these methods in visibility forecasting was confirmed through several parameters (i.e., root-mean-square error, mean absolute error, and mean absolute percentage error). The results show that: (1) among all meteorological parameters, wind speed was the best at reflecting the visibility change patterns; (2) long short-term memory recurrent neural networks (LSTM RNN), and gated recurrent unit (GRU) methods perform almost equally well on short-term visibility forecasts (i.e., 1, 3, and 6 h); (3) a classical machine learning method (i.e., the SVM) performs well in mid- and long-term visibility forecasts; (4) machine learning methods also have a certain degree of forecast accuracy even for long time periods (e.g., 7 2h).
Tema
visibility forecast; deep learning; machine learning; time-series forecasting
Idioma
eng
ISSN
ISSN electrónico: 2395-8812; ISSN impreso: 0187-6236

Enlaces