Data and analysis scripts for the paper:

Singh, M.S., Warren, R.A., & Jakob, C. (2019): A steady-state model for the relationship 
between humidity, instability, and precipitation in the tropics, Journal of Advances in 
Modeling Earth Systems (in review).

This archive includes three directories:

1) Figures
2) data
3) analysis

-----------------------------------------------------------------------------------------
1) Figures: This directory simply contains the figures within the paper.




-----------------------------------------------------------------------------------------
2) data: this directory includes model output and model configuration files for the
         simulations used in the publication. Some details of these simulations are 
         provided below, but a full description is given in the paper.

The data corresponds to output from a suite of simulations of what has been
termed "radiative-convective-dynamical equilibrium", in which convection within a limited
domain is simulated with a cloud-resolving model (in this case SAM) under the influence of
a constant large-scale flow. More specifically, we run SAM in a doubly periodic 
configuration over a fixed-SST boundary condition. The "large-scale flow" is implemented 
as a tendency on the prognostic fields in the model to mimic the effect of a domain-mean 
ascent or descent.

For our simulations, the large-scale ascent is characterised by a a sinusoidal shape with 
a maximum $wmax in the mid-troposphere reducing to zero at the tropopause. We run 
simulations for a variety of values of $wmax, but the large-scale flow is held constant 
for each simulation. 

Note that if $wmax is set to zero, the configuration corresponds to the well-known case of
radiative-convective equilibrium. In this case, our simulations match the configuration of
the recent RCEMIP (see: http://myweb.fsu.edu/awing/rcemip.html).

The tendencies induced by the large-scale flow may be applied in a number of ways. In our
default configuration (simple), the tendencies act only on the domain-mean temperature and
humidity fields, and horizontal advection is neglected. However, in alternate sets of 
simulations, we include tendencies on condensed water species and tendencies based on the 
total fields rather than the domain mean (full) or we include a representation of 
horizontal advection (HA). 

The simulations are run with SAM version 6.11.1 for 100 days, on a 96x96 km horizontal 
domain with 74 vertical levels and with a horizontal grid spacing of 1 km. We perform sets
of simulations at three SSTs (295, 300, 305 K) and with a range of values of $wmax ranging
from -0.6 cm/s to +0.5 cm/s. Full details of the simulations may be found in the 
aforementioned publication.

The data directory is organised in the following way:

Each simulation is contained within a directory with a naming convection as follows:

RCEMIP_SST$sst_$HA_w$wmax_ztop$ztop_96x96x74-1000m-8s_$simple

Here:

$sst       gives the SST in kelvin
$wmax      gives the maximum large-scale vertical velocity in m/s
$ztop      gives the height of the tropopause

$HA        is equal to 'HA' if horizontal advection is used, or ' ' if not.
$simple    is equal to '_simple' if the tendencies are applied to domain-mean 
	   humidity and temperature only, or ' ' if the tendencies are applied to the 
	   full fields and to condensed water.
           We also perform one simulation for which tendencies are applied to the full 
           temperature and humidity fields but NOT the condensed water fields; for this 
           simulation $simple = '_nocond'.
      
For the cases in which $wmax=0, only the SST is meaningful, and the naming system is 
simpler:

RCEMIP_SST$sst_96x96x74-1000m-8s
       

Within each simulation directory the file structure is as follows:

input/		        init_sounding.in	initial profile for simulation
			namelist_2400hr.in	namelist used to run model
			outputs_2400hr.in	file to specify output statistics
			qvprof.in		file to specify humidity outside domain
						for cases with advection ($HA = 'HA')
			vertical_grid.in	Vertical levels file
									

OUT_3D/		*810000.nc			A single snapshot of 3D variables at 75 days 

OUT_STAT/	summary*.nc			domain- and time-mean profiles 
						(and other statistics) from day 50-100.
						Included are profiles used to construct the Figures
						in the publication. In particular, statistics of 
						the vertical velocity distribution calculated 
						from the original 3D files are given.
             
             
             
-----------------------------------------------------------------------------------------
3) analysis: A set of analysis scripts that allows the figures to be created from the 
             archived data

THe analysis directory includes:

	- Plotting functions for each figure
	  "plot_FigN.m" (n = 1-7)

	- Script to calculate the equilibration timescale of simulations  
	  "calculate_equilibration_timescale.m" 

	- A matlab function which solves the plume model outlined in the paper. 
          "calculate_plume_solution.m"

	- The script used to archive the data from the original model output
	  "save_to_archive.m"

	- A package of file manipulation scripts used to read the model data
	  "+SAM" 

	- A package of thermodynamics scripts
	  "+atm" 
            
	- A package of scripts manipulating figures
	  "+fig"