Raw data from high-frequency observations of the ITB-MAWAS (Marine Automatic Weather Station) at Cirebon Jetty
Description
This dataset, titled “High-Frequency Observations dataset from the ITB Marine Automatic Weather Station (MAWAS) at Cirebon Jetty”, contains high-frequency coastal environmental observations collected at Cirebon Jetty, Indonesia (6.716131° S, 108.572581° E), from 1 October 2025 03:08:32 UTC to 4 May 2026 23:59:59 UTC. The quality-controlled release contains 3,298,281 records and includes the variables timestamp, water_level, temperature, humidity, pressure, wind_speed, and wind_direction. The dominant sampling interval is approximately 5 s, enabling analyses of short-timescale atmospheric and water-level variability. The dataset was compiled to provide an openly documented, high-frequency record for coastal monitoring, environmental analysis, and sensor-system evaluation, rather than to test a single hypothesis. Observations were acquired by the ITB MAWAS platform using RS485 sensors for wind speed, wind direction, air temperature, humidity, barometric pressure, and ultrasonic water-level measurement. Data were transmitted in real time through an ESP32-based logger and a cellular modem using MQTT to a remote server, then archived as daily JSON files. A key point for interpretation is the water_level variable. The ultrasonic sensor measured the air-column distance between the sensor and the dynamic water surface. Therefore, the published water-level series was transformed as 338 - raw water_level so that the released values represent corrected water level relative to the installation reference. Quality control was applied at the variable level using local spike screening based on rolling median and median absolute deviation, together with conservative range checks. Flagged outliers were not removed as rows; instead, affected values were replaced with blank entries while preserving the original timestamps. The dataset supports applications in coastal meteorology, nearshore monitoring, time-series analysis, data-quality assessment, and benchmarking of real-time marine observation systems in tropical coastal environments.
Files
Steps to reproduce
1. The ITB Marine Automatic Weather Station (MAWAS) was installed at Cirebon Jetty, Indonesia (6.716131° S, 108.572581° E). The station used RS485 sensors for wind speed (WS200 A-3-N), wind direction (WD300 A-3), air temperature-humidity-barometric pressure (S-THP-01A), and water level (750 cm ultrasonic level sensor). The system was powered by a 40 WP solar panel, solar charge controller, and 14 V lithium battery. 2. Sensor outputs were acquired by an ESP32-based logger. The logger transmitted observations in real time through a cellular modem using the MQTT protocol to a remote server. No local onboard storage was used; therefore, the archived data represent server-side records received from the telemetry stream. 3. Raw observations were archived as daily JSON files. Each record contained a UTC timestamp and the measured environmental variables. The official observation period used for this release spans from 1 October 2025 03:08:32 UTC to 4 May 2026 23:59:59 UTC. 4. The daily JSON files were merged into a unified time-series table. Only the variables used in the published dataset were retained: timestamp, water_level, temperature, humidity, pressure, wind_speed, and wind_direction. 5. Water level required transformation before publication. Because the ultrasonic sensor measured the air-column distance between the sensor and the dynamic water surface, the released water-level variable was calculated as: water_level = 338 - raw water_level. 6. Quality control was applied variable by variable using local spike detection based on rolling median and median absolute deviation, supported by conservative range checks. Values identified as outliers were replaced with blank entries, while timestamps were preserved. 7.The final release was exported as quality-controlled ASCII/CSV files for reuse in coastal meteorology, water-level analysis, and high-frequency environmental monitoring studies.
Institutions
- Bandung Institute of TechnologyWest Java, Bandung