National Solar Radiation Database (SUNY + TMY3)
This dataset works well for monthly and multi-year averages but performs poorly
for daily and hourly data when compared with medium-quality ground sites.
In recent years, thousands of land-based stations been deployed.
The difficulty was that no one had figured out how to
collect the data reliably or perform extensive quality control.
It says in the NSRD user manual that:
"Under partly cloudy skies,
because of the random and unknown location of the clouds, no model can
accurately estimate the solar radiation incident on the earth's surface at any
given time and location. Hence, the model used to estimate solar radiation when
measured data were not available was designed specifically to reproduce the
statistical and stochastic characteristics of multi-year solar radiation data
sets. This resulted in the sacrifice of accuracy for specific hours. Modeled
values for individual hours (under partly cloudy skies), therefore, may differ
greatly from measured values, had they been made." (
NSRDB User
Manual). It appears the primary purpose of this dataset is to provide
an accurate multi-year average based on the best data available. The best
data available 10 years ago was satellite data and a few ground sites. Now
there are thousands of high-quality and medium-quality ground sites. Why
not use them? We can demonstrate that medium-quality ground sites have
approximately half the error (rMAE) of satellite and modeled data in most
locations. High quality ground sites have even less error.
The NSRD manual continues, "It was anticipated that a multi year data base will
most often be used to create design- and typical-year subsets; to establish
normals, means, and extremes; and to select or evaluate sites for large solar
energy systems. Given these uses, it is important that simulated data sets
accurately represent the following statistical and stochastic characteristics of
measured data." NREL further admits
that 99% of this dataset was modeled and "should not be used to predict weather
for a particular period of time, nor is it an appropriate basis for evaluating
real-time energy production or efficiencies for building design applications or
solar conversion systems." (
TMY3 User Manual)
In other words, NREL says it's ok to use the data to select a site for a solar
power system, but it shouldn't be used for evaluating energy production or the
effiency.
There is clearly a need for high-quality hourly and daily measured data.
Private Modeling Companies
Our tougher competitors use scientific-quality ground measurements to correct
their satellite-based models. We can't say we're better at those locations
because we sometimes use the same data! The problem is that there are not
enough of these high quality sites to provide an accurate dataset for the entire
U.S. Their approach is to fill the gaps using satellite-based models. Our
approach is to fill the gaps using medium-quality ground sites. NREL has
recently published a manual showing relative accuracies of different data
sources, including the Solar Data Warehouse.
Check it out. For
more technical information, please read our paper from the 2011 ASES conference
in Raleigh:
Quality Analysis of Global Horizontal Irradiance Data from 3500 U.S.
Ground-Based Weather Stations by James Hall
PowerPoint
Presentation