Associated Issues

1. Negative temperature response from a BC pulse (https://github.com/JGCRI/hector/issues/127).
2. Temperature lag  from an emissions pulse (https://github.com/JGCRI/hector/issues/128).
Adria's almost ready to submit paper shows the response of Hector with doeclim to a series of impulse tests compared to other simple climate models. 

Description of changes

     DOECLIM

DOEClim is a fairly well known 1-D diffusive ocean model that is used in other simple climate models, Tanaka 2007,  cdice - https://github.com/scrim-network/cdice_doeclim, BRICK - https://github.com/scrim-network/BRICK. It consists of 3 main input parameters, equilibrium sensitivity, ocean heat diffusion and an aerosol scaling factor, from which it calculates a fast (mixed layer) and slow (deep ocean) temperature response.
  1. We removed the old temperature component  from Hector, rather than just disabling it.
  2. We removed the old heat flux calculation from the ocean component.  Now heat flux is calculated in the new temperature_component  
  3. The ini files for all four RCPs are updated and reflect the inputs to doeclim.
  4. Uncommented the volcanic forcing in the ini files.  
  5. Set diffusion value to 2.3 cm2/s which is also the value from the third IPCC assessment report and is what both MAGICC 5.3 and 6.0 use.