About

Since 1999 JHtech has partnered with many companies, from Fortune 500 to startups, in developing new technologies. These technologies are based on a convergence of principles from engineering, artificial intelligence and big data. Our successful projects include commercial solutions for stock trading, process control, market analysis and custom software. The principals hold advanced degrees in engineering and one holds a PhD in artificial intelligence, with over fifty years of combined experience. Each year billions of dollars of products are sold worldwide which incorporate our patents, technologies or product concept work.

Ten years ago, JHtech initiated a project to improve the global food supply through the application of artificial intelligence. Models were developed to predict daily crop growth and harvest yield. We amassed the nation’s largest database of crop yields and the corresponding climate at millions of locations, allowing us to identify which crop varieties are optimum for the various climates and soils across the United States. Professional services are provided to agricultural and the commodity markets through our companion website, HarvestDataWarehouse.com.

Our proprietary growth models predict a crop's day-by-day response to environmental factors: sunlight, soil, temperature and moisture. Plant growth and available moisture are very sensitive to solar radiation, and we found the models were more accurate when solar radiation measurements from nearby ground sites were utilized rather than satellite-based data. This led to the development of a significant new solar resource – our 2005-2018 U.S. Ground-Site Irradiance Database (GSID).

It is generally accepted that the solar radiation data from professional quality ground sites, when properly maintained, is more accurate than satellite estimates. In fact, it is recommended practice for major photovoltaic projects to validate long-term satellite data with one or more years of on-site measurements. Initially we were told that there were too few ground sites to compile a complete national database of measured solar radiation. Concerns were also expressed that the existing ground sites may not be well maintained, so the data might not be reliable. We believe we have addressed those concerns using statistical methods. Please see our white papers on the right side of this page for more information.

We have located over 7000 professionally operated ground sites measuring solar radiation across the United States. The average distance between sites in the continental US is about 40 miles - somewhat less dense in the great plains and denser in more populated areas. These sites belong to nearly 100 different federal, state and university networks. We have combined them into a “network of networks” so we can judge the accuracy of each individual site by comparing to nearby neighbors. Utilizing advanced filtering and data fusion techniques common in engineering and artificial intelligence, we carefully quality control the hourly and daily data from these 7000 ground sites into our database. Validation tests have shown that the daily and hourly GHI (global horizontal irradiance) measurements in our database are significantly more accurate than satellite-based estimates.

Using these same techniques, we have also completed similar solar radiation projects for other locations outside the US on a contract basis.

To our knowledge, the result of these efforts is the most comprehensive and accurate resource for US solar radiation currently available. We are making our daily GHI for 2005-2018 accessible free of charge via this website and invite potential users to evaluate it for their own applications. In addition, an atlas of monthly and annual total GHI is provided without charge. More detailed daily and hourly data from 2005 to the present, including professional analysis and services, is available on a custom project basis.

These images are a quick visual comparison of our continental US dataset with the National Solar Radiation Database made by NREL using satellite models. The contours match fairly well and we believe this demonstrates some potential benefits of combining ground and satellite datasets.

We welcome the opportunity to work with you – please contact us to discuss your specific needs.