Gis-based big data processing framework forlong-term forecasting of mountain river hydropower resources in multi-purpose hydropower complexes

Abstract

Relevance: Accurate long-term hydropower forecasting requires intelligent Big Data processing of heterogeneous hydrological and geospatial datasets.

Objective: This study develops a GIS-based Big Data processing framework for long-term forecasting of mountain river hydropower resources using distributed computing, statistical preprocessing, IoT monitoring, and geospatial integration technologies to improve forecasting accuracy and operational reliability of multi-purpose hydropower complexes.

Methods: PySpark-based distributed processing, Gaussian filtering, log-normal transformation, Poisson and binomial statistical preprocessing methods were applied.

Results: The proposed framework improved data reliability to 0,96–0,99, reduced anomalies, and enhanced long-term hydropower forecasting stability and processing efficiency.