You need to agree to share your contact information to access this dataset

This repository is publicly accessible, but you have to accept the conditions to access its files and content.

Log in or Sign Up to review the conditions and access this dataset content.

Panj River Hydrological Dataset (1970–1990)

Daily hydrological and meteorological observations at the Khirmandzho gauging station (Panj River, Tajikistan) for the period 1970–1990.

Data Sources

  • Hydrological data: Official publications of the "Hydrological Daily Journal. Basin of Central Asian Rivers. Amu Darya and Zarafshan Basins" (Khirmandzho gauging station).

  • Meteorological data: Dushanbe station (ID 38836) from the open archive Meteostat.

  • Processing: Merging, interpolation of missing values, addition of lags and rolling statistics (available in supplementary files – see below).

  • The data preprocessing and modeling were performed using the ML Pipeline PRO Suite software (State Registration Certificate No. 2026615844).

Dataset Structure

The main file panj_river_daily.parquet contains the following columns:

Column Description
tavg Average daily air temperature (°C)
tmin Minimum daily air temperature (°C)
tmax Maximum daily air temperature (°C)
raskhodvoda River discharge (m³/s)
urovenvoda Water level (arbitrary units)
prcp Daily precipitation total (mm)

Frequency: Daily
Period: 1970-01-01 – 1990-12-31
Number of records: 7670

Basic Statistics

Statistic tavg tmin tmax raskhodvoda urovenvoda prcp
count 7670.00 7670.00 7670.00 7670.00 7670.00 7602.00
mean 15.00 8.44 22.43 774.52 367.73 1.68
std 9.45 7.96 10.80 636.88 134.45 5.02
min -13.80 -22.30 -7.70 22.75 69.75 0.00
25% 7.40 2.30 13.70 301.98 258.85 0.00
50% 15.60 9.00 23.50 442.57 331.36 0.00
75% 23.30 15.00 32.00 1313.62 485.32 0.00
max 33.90 25.40 49.00 2657.71 845.48 81.90

Missing Values

Column Missing (%)
prcp 0.89

Sample Data

import pandas as pd

df = pd.read_parquet("hf://datasets/TimeSeriesHub/panj-river-daily-1970-1990/data/panj_river_daily.parquet")
print(df.head())
date tavg tmin tmax raskhodvoda urovenvoda prcp
1970-01-01 7.8 5.5 11.4 336.23 211.06 12.5
1970-01-02 5.4 0.1 10.9 341.28 204.11 20.8
1970-01-03 -0.3 -2.3 3.4 347.45 196.64 6.8
1970-01-04 0.0 -5.1 6.6 346.27 196.15 0.0
1970-01-05 1.9 -4.3 6.9 344.40 196.07 0.0

Usage Examples

The dataset can be used for:

  • Hydrological modeling and streamflow forecasting.
  • Climate change analysis in the Panj River basin.
  • Training machine learning models (regression, time series).
  • Supporting engineering design of hydraulic structures (dams, bank protection).

Loading the data with Python

import pandas as pd

df = pd.read_parquet("hf://datasets/TimeSeriesHub/panj-river-daily-1970-1990/data/panj_river_daily.parquet")
# 'date' is the index; if you need it as a column:
df = df.reset_index()

Loading with R

library(arrow)
df <- read_parquet("hf://datasets/TimeSeriesHub/panj-river-daily-1970-1990/data/panj_river_daily.parquet")

License

The data are provided for scientific and educational purposes. When using this dataset, please cite the original sources and this repository.

Citation

If you use this dataset in your research, please cite it as:

Arabov, M.K., Fathulloev, N.I.(2025). Panj River Hydrological Dataset (1970–1990) [Data set]. Hugging Face. https://huggingface.co/datasets/TimeSeriesHub/panj-river-daily-1970-1990

Related Publications

If you use this dataset in your research, please also consider citing the following papers that describe the methodology and applications:

  1. Fathulloev, N.I., Arabov, M.K. (2025). Ensembling Machine Learning Models to Enhance the Accuracy of Extreme Water Discharge Forecasting Considering Hydrometeorological Characteristics. Science and Innovation. Series of Geological and Technical Sciences, Tajik National University, No. 4, pp. 136–150. (in Russian)

  2. Arabov, M.K., Fathulloev, N.I., Nurov, I.Zh. (2025). Forecast of Daily Water Discharge for the Design of Shore Protection Structures Based on a Hybrid Fuzzy Time Series Model. Proceedings of the International Conference “Scientific research of the SCO countries: synergy and integration”, Beijing, PRC, 30 Dec. 2025, pp. 122–129. DOI: 10.34660/INF.2025.73.25.059

  3. Arabov, M.K., Fathulloev, N.I. (2026). ML Pipeline PRO Suite [Computer software]. Russian Federation State Registration Certificate No. 2026615844. (Software used for data processing and modeling)

Contact

For questions or suggestions, please contact the dataset authors:

Dataset creation date: 2026-03-19

Downloads last month
9