Dataset of ground-based meteorological observations and sunshine duration for solar irradiance modelling in Central Java, Indonesia
Description
This dataset provides three years (1 January 2022 - 31 December 2024) of daily ground-based meteorological observations from the Sempu climatological station in Central Java, Indonesia (7.411 S, 110.237 E), together with a co-located NASA POWER daily time series retrieved for the same coordinates and period, and a quality-controlled, analysis-ready file merging the two sources. The ground-based observations are distributed by the Central Java Provincial Public Works Agency for Water Resources and Spatial Planning (Dinas PUSDATARU). They comprise daily mean air temperature, relative humidity, wind speed, and sunshine duration (as a percentage of the maximum possible duration). The NASA POWER series provides all-sky surface shortwave downward irradiance (global horizontal irradiance, GHI) and ancillary 2 m meteorology. Long, openly licensed, station-level daily records from the Indonesian maritime continent are scarce. By pairing measured sunshine duration with a co-located satellite/reanalysis series on a common daily time base, the dataset supports developing and evaluating sunshine-based empirical estimates of daily GHI under humid tropical conditions, benchmarking machine-learning irradiance models, and cross-comparing ground meteorology against reanalysis. It also suits reference-evapotranspiration, agro-climatic, and renewable-energy resource assessment. Notes for reuse. (1) GHI is a NASA POWER satellite-derived product, not a ground pyranometer measurement; the station has no pyranometer. The data support estimating GHI from ground meteorology, not validating satellite irradiance against an independent ground reference. (2) The source reports daily maximum and minimum temperature, but on about 82% of days the daily range is exactly 1.0 C and the mean is their exact midpoint, indicating the extremes are derived from the mean rather than measured. They are kept only in the raw file for transparency and excluded from the analysis-ready file, which retains the daily mean temperature (Tavg). Files. (a) sempu_klimatologi_2022_2024.xlsx - raw observations, 1,096 records (date, RH, Tmax, Tmin, Tavg, WS, Sunshine_%). (b) Data_power_nasa_sempu.xlsx - NASA POWER, 1,096 records (Date, GHI, T2M, RH2M, WS2M). (c) master_dataset_irradiance_sempu.csv - analysis-ready, 1,056 records (date, Tavg, RH, WS, Sunshine_%, GHI). (d) qc_merge_sempu.py - script reproducing (c). (e) README.txt - documentation. Units: Tavg C; RH %; WS km/h; Sunshine_% %; GHI kWh/m2/day. Dates local time (WIB, UTC+7), ISO 8601; point decimal separator; empty cells = missing.
Files
Steps to reproduce
The analysis-ready file was produced from the two raw files with the included script qc_merge_sempu.py (Python 3, requires pandas, numpy, openpyxl). Quality control: (1) the raw station file was reindexed to the full 2022–2024 daily calendar (1,096 days; 34 days have no station record); (2) a physical-range test was applied to the retained variables (Tavg 15–35 °C, RH 0–100 %, WS ≥ 0, Sunshine_% 0–100 %), which flagged four days with corrupt temperature (58–63 °C, coinciding with impossible minimum-temperature entries of 91–99 °C) and set them to missing; (3) daily maximum and minimum temperature were excluded because they are derived from the daily mean rather than measured; (4) the station file was merged with the NASA POWER series on date (inner join); (5) days with any missing or flagged value were removed by listwise deletion, yielding 1,056 complete daily records. Constant runs longer than three days occur in relative humidity and mean temperature and are reported rather than removed. Run: python qc_merge_sempu.py in the folder containing the two raw files.
Institutions
- Diponegoro UniversityCentral Java, Semarang