Annually and monthly resolved solar irradiance and atmospheric temperature data across the Hawaiian archipelago from 1998 to 2015 with interannual summary statistics

This article contains data and summary statistics of solar irradiance and dry bulb temperature across the Hawaiian archipelago resolved on a monthly basis and spanning years 1998–2015. This data was derived in association with an article titled “Consequences of Neglecting the Interannual Variability of the Solar Resource: A Case Study of Photovoltaic Power Among the Hawaiian Islands” (Bryce et al., 2018 [7]). The solar irradiance data is presented in terms of Direct Normal Irradiance (DNI), Diffuse Horizontal Irradiance (DHI), and Global Horizontal Irradiance (GHI) and was obtained from the satellite-derived data contained in the National Solar Radiation Database (NSRDB). The temperature data is also obtained from this source. We have processed the NSRDB data and compiled these monthly resolved data sets, along with interannual summary statistics including the interannual coefficient of variability.


Data
The U.S. Department of Energy has funded the multiyear production of the National Solar Radiation Database (NSRDB). This database has been updated three times since the initial public release in 1992. The most recent version contains a wide range of solar irradiance data derived from satellite data and meteorological data derived from reanalysis data. This data is presented with a 4-km by 4-km spatial resolution and half-hourly temporal resolution covering the 18 years from 1998 to 2015 [2]. The reanalysis data is obtained from NASA's Modern-Era Retrospective Analysis for Research and Applications (MERRA) data set [3]. This version of the NSRDB includes data from the area bounded by the 25°W and 175°W meridians and by the À 20°S and 60°N parallels.
The performance of the irradiance data sets in the most recent version of the NSRDB was evaluated in [2]; the hourly average satellite-derived data have a mean bias error of 7 5% for GHI and less than 7 10% for DNI when compared against concurrent ground-based measurements from seven sites of the Surface Radiation Budget Network. The NSRDB has been shown to accurately feature interannual variability; the coefficient of variation (COV) for both the satellite-derived data sets and ground-measured from seven Surface Radiation Budget Network stations across the United States where compared. The data sets show similar annual trends and differ by only 0.68% COV for global horizontal irradiance (GHI) and 0.97% COV for direct normal irradiance (DNI).
In the original public release, [3] reported the root mean squared residuals between the MERRA data sets and measured data. The MERRA data sets have also been independently evaluated in [4][5][6]; the data sets feature a mean bias of dry-bulb temperature of less than 1°C globally.   Herein, we share the monthly averaged solar irradiance and dry-bulb temperature data across all eight major islands of the Hawaiian Archipelago for all years from 1998 to 2015. We report box-plots of values from all the NSRDB grid cells, as well as tabulated island-wide mean, for each island. Along with the tabulated data, we report interannual summary statistics to quantify the interannual variability of solar irradiance, reported as GHI, and temperature for each island. This data and the summary statistics were compiled alongside another work which aimed to characterize the interannual variability of the solar irradiance and meteorological conditions across the State of Hawaii and to then characterize the variability of PV power generation [7].
Solar irradiance values are often reported in a way that excludes values below a certain threshold of the clear-sky radiation [8]. This filter excludes the shading effect that is pronounced during dawn and dusk. To identify the appropriate threshold, we calculated the first nonzero irradiance value during each day for each site across the State of Hawaii; the mean value for GHI and DNI was 65 W/ m 2 , whereas a lower value of 35 W/m 2 for DHI was identified. No filter is applied to the temperature data. For all NSRDB cells across each island, the annual mean value was calculated. Then, a box plot of the annual mean value of all NSRDB cells was produced for each year of data . The data that was used to generate these box plots, including the minimum values, the values of the first quartiles, the values of the second quartiles, the values of the third quartiles, and the maximum annual mean values are tabulated in Tables 1-4.
The grand mean values of GHI and temperature were calculated among all NSRDB cells on both a monthly and annual basis. Additionally, the monthly grand means were compared on an interannual basis. The 1st, and 3rd quartiles of the 18 years of monthly grand means were calculated. On the same interannual basis, the mean, the standard deviation, the maximum, the minimum, and range of monthly grand means were calculated. Finally, the interannual coefficient of variation (COV), which is the standard deviation divided by the mean and expressed as a percentage, was calculated (see Tables 5-20).  Table 1  The minimum values, the quartile values, and the maximum annual mean values for DNI across each island during each year from 1998 to 2015.   1998  1999  2000  2001  2002  2003  2004  2005  2006  2007  2008  2009  2010  2011  2012  2013  2014 6 Q2 191.3 197.8 189.5 203.7 200.6 194.5 198.5 188.3 196.5 192.8 192.0 197.8 191.4 201.9 178 . 1 190.2 192.9 192.4 202.4 192.6 199.5 190.6 195.5 196.1 197.1 198 Table 9 Monthly averaged global horizontal irradiance (W/m 2 ) during daylight hours across Maui from 1998 to 2015 with annual means and interannual statistics.     Table 11 Monthly averaged global horizontal irradiance (W/m 2 ) during daylight hours across Ni'ihau from 1998 to 2015 with annual means and interannual statistics. 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013