A Daily Dataset of Solar Radiation, Reference Evapotranspiration (ET0), and Potential Evapotranspiration (PET) Derived from Bangladesh Meteorological Department Climate Observations
Description
This dataset provides a standardized, quality-controlled, analysis-ready daily record of solar radiation (Rs), reference evapotranspiration (ET0), and potential evapotranspiration (PET) for Bangladesh, covering 1 January 1961 to 31 December 2023 (63 years). It comprises 528,676 daily records from 35 synoptic meteorological stations operated by the Bangladesh Meteorological Department (BMD), distributed across the eight administrative divisions of Bangladesh. Solar radiation was estimated from observed sunshine duration using the Ångström–Prescott equation (FAO-56). Reference evapotranspiration (ET0) was calculated using the FAO-56 Penman–Monteith method, and potential evapotranspiration (PET) was derived using the standard Penman approach (McMahon et al., 2013). All records underwent quality-control procedures including range validation, internal consistency checks, missing-data screening, and temporal homogeneity verification. Each record includes station identification, geographic coordinates (latitude, longitude, elevation), the observation date (with day/month/year and derived season/decade fields), and the three derived agro-climatic variables (Rs, ET0, PET). The dataset supports research in climate change assessment, irrigation water management, hydrological and crop-simulation modeling, drought monitoring (e.g., SPEI), and machine-learning applications for evapotranspiration prediction across Bangladesh and the broader South Asian region. Files included: the full dataset (CSV and XLSX), a data dictionary describing all variables, and a station metadata table.
Files
Steps to reproduce
WITH base AS ( SELECT s.station_id, s.date_day, s.date_month, s.date_year, s.max_temp, s.min_temp, s.rh, s.wind_speed, s.sunshine, s.altitude, RADIANS(s.latitude) AS latitude_rad, (s.max_temp + s.min_temp) / 2 AS t_mean, DAYOFYEAR(STR_TO_DATE( CONCAT(s.date_year, '-', s.date_month, '-', s.date_day), '%Y-%m-%d' )) AS doy -- FIXED: real day-of-year FROM station_climate_data s WHERE s.date_year = 2023 ), astro AS ( SELECT b.*, 0.409 * SIN(2*PI()/365 * b.doy - 1.39) AS delta, 1 + 0.033 * COS(2*PI()/365 * b.doy) AS dr FROM base b ), daylength AS ( SELECT a.*, ACOS(-TAN(a.latitude_rad) * TAN(a.delta)) AS omega_s FROM astro a ), radiation AS ( SELECT d.*, (24*60/PI()) * 0.0820 * d.dr * (d.omega_s * SIN(d.latitude_rad) * SIN(d.delta) + COS(d.latitude_rad) * COS(d.delta) * SIN(d.omega_s)) AS Ra, 24/PI() * d.omega_s AS N, 0.6108 * EXP(17.27 * d.t_mean / (d.t_mean + 237.3)) AS e_s, 0.6108 * EXP(17.27 * d.min_temp / (d.min_temp + 237.3)) AS e_s_tmin, 4098 * (0.6108 * EXP(17.27 * d.t_mean / (d.t_mean + 237.3))) / POWER(d.t_mean + 237.3, 2) AS delta_svp, 101.3 * POWER((293 - 0.0065 * d.altitude) / 293, 5.26) AS P FROM daylength d ), psychro AS ( SELECT r.*, 0.000665 * r.P AS gamma, r.e_s_tmin * r.rh / 100 AS e_a FROM radiation r ), net_radiation AS ( SELECT p.*, (1 - 0.23) * p.Ra * (0.25 + 0.5 * (p.sunshine / p.N)) AS R_ns, 4.903e-9 * POWER(p.t_mean + 273.16, 4) * (0.34 - 0.14 * SQRT(p.e_a)) * (1.35 * (0.25 + 0.5 * (p.sunshine / p.N)) - 0.35) AS R_nl FROM psychro p ), pm_core AS ( SELECT n.*, ROUND(n.R_ns, 2) AS solar_radiation, (0.408 * n.delta_svp * (n.R_ns - n.R_nl) + n.gamma * (900 / (n.t_mean + 273)) * n.wind_speed * (n.e_s - n.e_a)) / (n.delta_svp + n.gamma * (1 + 0.34 * n.wind_speed)) AS pm_value FROM net_radiation n ) SELECT station_id, date_day AS day, date_month AS month, date_year AS year, max_temp, min_temp, rh, wind_speed, sunshine, solar_radiation, ROUND(pm_value, 2) AS eto, ROUND(pm_value, 2) AS pet, -- still a placeholder, see note above 'BMD' AS data_source FROM pm_core ORDER BY station_id, month, day;
Institutions
- Bangladesh Agricultural Research CouncilDhaka Division, Dhaka
- University of Frontier Technology, BangladeshDhaka Division, Gazipur