AN INTERNATIONAL SOLAR IRRADIANCE DATA INGEST SYSTEM FOR FORECASTING SOLAR POWER AND AGRICULTURAL CROP YIELDS

摘要

A data system has been developed for the real-time ingest of solar irradiance and climate data from thousands of ground-based measurement sites. Compared to space-based sensing of solar radiation, ground sensors can offer advantages in terms of accuracy and latency. The system is currently focused on the United States and the European Union, with data from other countries to be included in the future. It currently provides historical data for accurate solar resource assessment, real-time data for solar power forecast models, and data inputs for agricultural crop yield models. This ingest is unique in terms of the number of sites providing irradiance data, and the international scope of the project. Outputs include continuous gridded and point global horizontal irradiance data for the day, continuous irradiance forecasts for the next four hours, and historical summaries of solar and climate data. The system archives the observations and includes routines to back-fill data that could not be obtained in real-time. Additionally the system monitors each sensor by analyzing the time history of observations, comparing measurements from nearby sites, and checking observations against theoretical solar envelopes. Constructing a real-time ingest of ground-based data presented unique challenges such as asynchronous data availability, missing or delayed data, uncertainty about sensor maintenance and calibration, and the need to access many different networks with different data formats. This paper describes the methodology for addressing each of these problems and the current system performance.