Initial commit of historical leaf temp/par averaging approach.

[drnimbusrain]

@MaggieMarvin This is a first crack at explicitly calculating historical leaf temp and par values used in canopy-app. I have created a new option, "hist_opt" to turn this on (hist_opt=1). See the description in the README, but if the canopy-app simulation is < 24 hours then instantaneous values will still be used, and if turned on and run > 24 hours, historical running avearges can be used. Can you closely test this using a longer simulation and analyzing the output after 24 hours against MEGAN and PECORINO?

This is currently a draft and needs thorough testing:

• Compare spatial impacts of historical averaging against MEGANv2 and MEGANv3
• Test historical approach for > 24 hours for assessing the e-folding running averages
• Test historical + gamma_sm impacts.

[drnimbusrain]

@MaggieMarvin I have merged develop (with gamma_sm changes) into historical for your testing of both options on and running it longer to test the running average feature when > 24 hours.

[MaggieMarvin]

Looks good! The moving 24-hour average behaves as expected, for example different length simulation windows produce equivalent results for the same time and no historical effects are applied before 25 hours. For our test case of 08/01/2021 1800 UTC, turning on historical effects has a significant impact, reducing emissions across the region by about 40% on average (see attached plot), and improving agreement with MEGANv3 approaching a factor of 2.