Deep-learning-based prediction of precipitable water vapor in the Chajnantor area

¹ Facultad de Ingeniería, Universidad Católica de la Santísima Concepción, Alonso de Ribera 2850, Concepción, Chile
² Departamento Ingeniería Informática y Ciencias de la Computación, Universidad de Concepción, Concepción 4070409, Chile
³ Facultad de Ingeniería, Universidad del Bío-Bío, Collao 1202, Concepción, Chile
⁴ Departamento de Electrónica e Informática, Universidad Técnica Federico Santa María - Sede Concepción, Arteaga Alemparte 943, Concepción, Chile
⁵ Departamento de Ingeniería Eléctrica, Universidad Católica de la Santísima Concepción, Alonso de Ribera 2850, Concepción, Chile
⁶ Centro de Energía, Universidad Católica de la Santísima Concepción, Alonso de Ribera 2850, Concepción, Chile
⁷ National Astronomical Observatories, Chinese Academy of Sciences, Beijing 100101, China
⁸ Nanjing Institute of Astronomical Optics & Technology, Chinese Academy of Science, Nanjing 210042, China
⁹ CAS Key Laboratory of Astronomical Optics & Technology, Nanjing Institute of Astronomical Optics & Technology, Nanjing 210042, China
¹⁰ Chinese Academy of Sciences South America Center for Astronomy, National Astronomical Observatories, CAS, Beijing 100101, China
¹¹ Instituto de Astronomía, Universidad Católica del Norte, Av. Angamos 0610, Antofagasta, Chile
¹² CePIA, Departamento de Astronomía, Universidad de Concepción, Casilla 160 C, Concepción, Chile

^★ Corresponding author: This email address is being protected from spambots. You need JavaScript enabled to view it.

Received: 25 June 2025
Accepted: 22 November 2025

Abstract

Context. Astronomical observations at millimeter and submillimeter wavelengths strongly depend on the amount of precipitable water vapor (PWV) in the atmosphere, which directly affects the sky transparency and decreases the signal-to-noise ratio of the signals received by radio telescopes.

Aims. Predictions of PWV at different forecasting horizons are crucial to supporting the telescope operations, engineering planning, scheduling and observational efficiency of radio observatories installed in the Chajnantor area in northern Chile.

Methods. We developed and validated a long short-term memory (LSTM) deep-learning-based model to predict PWV at forecasting horizons of 12, 24, 36, and 48 hours using historical data from two 183 GHz radiometers and a weather station in the Chajnantor area.

Results. We find that the LSTM method is able to predict PWV in the 12- and 24-hour forecasting horizons with a mean absolute percentage error of ~22% compared to ~36% for the traditional Global Forecast System method used by the Atacama Pathfinder Experiment, and the root mean square error is reduced by ~50%.

Conclusions. We present a first application of deep learning techniques for preliminary predictions of PWV in the Chajnantor area. The prediction performance shows significant improvements compared to traditional methods in 12- and 24-hour time windows. We also propose strategies to improve our method on shorter (<12 hour) and longer (>36 hour) forecasting timescales.