dli_details.md

Daily light integral (DLI)

The DLI is all the photosynthetically active radiation (PAR) at a location on the earth's surface, summed from sunrise until sunset.

The units are moles of photons per square meter per day --- mol m^-2 d^-1.

Where these data come from

The NASA POWER agroclimatology dataset provides satellite data of many useful solar and meteorological parameters on a 0.5° x 0.5° grid, including all sky insolation incident on a horizontal surface, which is global solar radiation (R_s). On a daily basis, the units for R_s are megajoules per square meter per day, MJ m^-2 d^-1.

There is a consistent proportion of PAR in R_s. I've often used a conversion factor of 2.04 (multiplying R_s by 2.04 to get the DLI in mol m^-2 d^-1) based on Meek et al. (1984). In this app, I've used the same conversion as was used by Faust and Logan (2018), "which assumes that 45% of the solar radiation is in the photosynthetically active radiation (PAR, 400–700 nm) region and 4.48 μmol J⁻¹ is the conversion from radiometric to quantum units." That works out to be a 2.01 conversion factor, rather than 2.04. Basically, you can multiply R_s by 2 and be really close to the DLI.

The data I show in this app as DLI are the NASA POWER all sky insolation incident on a horizontal surface (R_s) data multiplied by 0.45 and then by 4.48.

How I got the data

I used the nasapower package (in R software) to obtain the data. The nasapower package is maintained by Adam Sparks.

This is a Shiny app and I used the leaflet package for the map in the app.

Any missing data have been approximated using the na.approx function from the zoo package. I haven't seen a lot of missing values but that is run after the data are downloaded. The NASA POWER global solar radiation data are usually available up to about 5 days prior to the current date. This app is getting data from 370 days ago until 6 days ago, which is a total of 364 days, and is exactly 52 weeks. Those 364 days (52 weeks) of data are shown on the charts.

There are different ways to deal with weeks, and what I have done in this app is calculate them as 7 days periods that begin with the first day in the 364 days of data shown.

For more information

• The code for this app is at https://github.com/micahwoods/global_dli.

• Please file any bug reports or feature requests at https://github.com/micahwoods/global_dli/issues.

• Posts tagged light on the ATC website

• NASA Power

• Instantaneous PAR app based on time of day and location. I'm going to update that one sometime to make it easier to input your location. Like with a map.

• Asian Turfgrass Center