Skip to content

E3SM Single Column Model Case Library

bogensch edited this page Jul 22, 2019 · 67 revisions

Directory

Verified Cases:

AEROSOLINDIRECT - Study of Aerosol Indirect Effects in China

ARM95 - Deep Cumulus Convection

ARM97 - Deep Cumulus Convection

ARM GCSS Shallow Convection - Diurnal Cycle of Continental Shallow Convection

ARM - SGP Continuous Forcing (2004 - 2015)

BOMEX - Trade Wind Cumulus

DYCOMS RF01 - Nonprecipitating Marine Stratocumulus

DYCOMS RF02 - Drizzling Subtropical Stratocumulus

DYNAMO AMIE - Dynamics of the Madden Julian Oscillation

DYNAMO North Sounding - Dynamics of the Madden Julian Oscillation

DYNAMO Revelle - Dynamics of the Madden Julian Oscillation

GATE - Tropical Atlantic Deep Convection

GOAMAZON - Green Ocean Amazon

ISDAC - Indirect and semi-direct aerosol campaign

MC3E - Mid-latitude Continental Convective Clouds Experiment

MPACE - Mixed-Phase Arctic Cloud Experiment

MPACE-B Mixed-Phase Stratocumulus Intercomparison

RICO - Trade-Wind Cumulus

TOGA-COARE - Tropical Pacific Warm Pool Deep Convection

TWP-ICE Tropical Pacific Warm Pool Deep Convection

Unverified Cases:

ARM SGP - Southern Great Plains Site

ATEX - Trade Wind Cumuli under a Strong Inversion

DARWIN - Forcing from tropical western pacific

RACORO - Continental Boundary Layer Clouds

SPARTICUS - Cirrus

E3SM Replay Mode

REPLAY - Replay a column from an E3SM run

Background

The goal of this page is to serve as a reference for running the E3SM SCM, in addition to serving as the library for the SCM cases. For each case below, a brief orientation to the case is given along with a description of why you may want to use it. Links are then provided to an IOP file containing the input data needed to run the case along with a sample run script which configures, compiles, and runs that case. For select verified cases, output from the simulations we did which convinced us the simulation was performed correctly are also included for use as a sanity check that your implementation is performing as intended.

NOTE that recent E3SM master (post May 11, 2017, commit: 43c24616cf3ce0d44d680ab816f0ad5e37e4ec67) should be used to run the E3SM SCM. If not using E3SM git hooks, issue a 'git submodule update --init' when you update to the latest hash of master to get the new submodule (FATES).

Verified Cases

We have checked that the run script and IOP files for the cases below are free of obvious bugs and design flaws when run in E3SM.

AEROSOLINDIRECT - Study of Aerosol Indirect Effects in China

Overview: These forcings are derived from observations taken as part of the Study of Aerosol Indirect Effects in China (AEROSOLINDIRECT). The ARM mobile facility was deployed to study the Yangtze delta region in southeast China, the center of the satellite rainfall retrieval anomaly and also the area where the aerosol optical depth distribution in China is maximal. This is a 29 day case spanning November 2008.

AEROSOLINDIRECT SCM Run Script

ARM95 - Deep Cumulus Convection

Overview: This case is based on the IOP that took place in the Southern Great Plains (SGP) in July 1995, which covered a total of 30 days. This is a convectively driven case Zhang and Lin (1997) with a period that contained a wide range of summertime weather conditions. The first segment of this case is dominated by local convection and frequent heavy precipitation, while the second segment was generally clear and hot. The last segment of the case was affected by a large, convective complex with sustained precipitation. Data from this IOP formed the basis for the ARM/GCSS Case 1 SCM intercomparison by Ghan et al. (2000). The forcing data was developed using the constrained variational analysis method described in Zhang and Lin (1997) and Zhang et al. (2001).

Verification Notes: This case is simulated with the expected behavior, with dry periods and segments of vigorous deep convection.

ARM95 SCM Run Script

ARM97 - Deep Cumulus Convection

Overview: This IOP occurred at the same SGP site as the ARM95 case in June and July of 1997. Similar to the ARM95 case, this case also features several distinct periods characterized by a wide range of summertime weather conditions. Data from this IOP formed the basis for the ARM/GCSS case 3 SCM/CRM intercomparison by Xie et al. (2002) and Xu et al. (2002). The forcing data was developed using the constrained variational analysis method described in Zhang and Lin (1997) and Zhang et al. (2001).

Verification Notes: This case is simulated with the expected behavior, with dry periods and segments of vigorous deep convection.

ARM97 SCM Run Script

ARM GCSS - Diurnal Cycle of Continental Shallow Convection

Overview: This is a case that simulates the development of shallow cumulus convection over land at the ARM SGP site (Brown et al. 2002) from 21 June 1997. This is a good test case of shallow convection that is an unsteady case and more strong forced, compared to the marine shallow convective cases. The GCSS set-up for this case was run with no microphysics and this SCM run is configured to replicate that idealization.

ARM GCSS Run Script

ARM - SGP Continuous Forcing (2004 - 2015)

Overview: This case provides continuous forcing from the ARM Southern Great Plains (SGP) site from January 2004 through December 2015. It is possible to run this case for the entire 11 year period, but users are encouraged to select a desired subset time period to run on (options provided in the script).

ARM SGP Continuous Forcing Run Script

BOMEX - Trade Wind Cumulus

Overview: This case is based on the Barbados Oceanographic and Meteorological Experiment (BOMEX) and represents a case of undisturbed trade wind cumulus under steady-state conditions. The forcing for this case follows that of the LES intercomparison study of Siebesma et al. (2003). The GCSS set-up for this case was run with no microphysics and this SCM run is configured to replicate that idealization.

BOMEX Run Script

DYCOMS RF01 - Nonprecipitating Marine Stratocumulus

Overview: Research flight 1 of the second Dynamics and Chemistry of Marine Stratocumulus field campaign (hereafter DYCOMS RF01) sampled sampled nonprecipitating marine stratocumlus under a strong inversion. the forcing for this case follows that of the LES intercomparison study of Stevens et al. (2005). The GCSS set-up for this case was run with no microphysics and this SCM run is configured to replicate that idealization.

DYCOMS RF01 Run Script

DYCOMS RF02 - Drizzling Subtropical Stratocumulus

Overview: Research flight 2 of the second Dynamics and Chemistry of Marine Stratocumulus field campaign (hereafter DYCOMS RF02) sampled drizzling stratocumulus off the coast of California during the night of 11 July 1999. Data from this flight formed the basis for a SCM intercomparison by Wyant et al. (2007) and an LES intercomparison by Ackerman et al. (2009). This is a classic test case for microphysical processes in stratocumulus, which are a cloud regime that climate models traditionally struggle with and which are critical to the planetary energy budget.

Verification Notes: Simulation is initialized with a stratocumulus layer that is maintained once simulation spins up.

DYCOMS RF02 Run Script

DYNAMO AMIE - Dynamics of the Madden Julian Oscillation

Overview This is a 90 day case from the Dynamics of the Madden Julian Oscillation (DYNAMO) field campaign which was an IOP over the Indian Ocean to collect data to study the onset of the MJO. This particular case includes forcing data collected as part of the AMIE-Gan field campaign.

DYNAMO AMIE Run Script

DYNAMO North Sounding - Dynamics of the Madden Julian Oscillation

Overview This is a 90 day case from the Dynamics of the Madden Julian Oscillation (DYNAMO) field campaign which was an IOP over the Indian Ocean to collect data to study the onset of the MJO. This particular case includes forcing data collected from the north sounding array.

DYNAMO North Sounding Run Script

DYNAMO Revelle - Dynamics of the Madden Julian Oscillation

Overview This is a 90 day case from the Dynamics of the Madden Julian Oscillation (DYNAMO) field campaign which was an IOP over the Indian Ocean to collect data to study the onset of the MJO. This particular case includes forcing data collected aboard the R/V Roger Revelle operated by the Scripps Institution of Oceanography.

DYNAMO Revelle Run Script

GATE - Tropical Atlantic Deep Convection

Overview: This case is based on the Global Atmospheric Research Program's Atlantic Tropical Experiment (GATE, Houze and Betts 1981), who's goal was to improve basic understanding of topical convection and its role in the global atmospheric circulation. This is a 20 day case that begins on 30 August 1974.

Verification Notes: This case is simulated with the expected behavior.

GATE Run Script

GOAMAZON - Green Ocean Amazon

Overview This is a 23 month case from the Observations and Modeling of the Green Ocean Amazon (GOAMAZON) field campaign (Martin et al. 2017). Observations and forcing are derived from the ARM Mobile Facility, which was located downwind of the city of Manaus, Brazil near Manacapuru from January 2014 to November 2015. The site was situated so that it experienced the extremes of (i) a pristine atmosphere when the Manaus pollution plume meandered and (ii) heavy pollution and the interactions of that pollution with the natural environment when the plume regularly intersected the site.

GOAMAZON Run Script

ISDAC - Indirect and semi-direct aerosol campaign

Overview: The Indirect and semi-direct aerosol campaign is an intensive cloud and aerosol IOP conducted at ARM's North Slope of Alaska (NSA) locale for three weeks in April 2008. This is a 29 day case.

ISDAC Run Script

MC3E - Mid-latitude Continental Convective Clouds Experiment

Overview: The Midlatitude Continental Convective Clouds Experiment took place from April 22 - June 6, 2011, near Lamont, Oklahoma in the region surrounding the Department of Energy (DOE) Atmospheric Radiation Measurement (ARM) Program Southern Great Plains Central Facility. A broad variety of cloud and precipitation types were sampled during the MC3E including cumulus clouds, severe thunderstorms, mesoscale convective systems and widespread regions of stratiform precipitation.

MC3E Run Script

MPACE - Mixed-Phase Arctic Cloud Experiment

Overview: The Mixed-Phase Arctic Cloud Experiment (MPACE, Verlinde et al. 2007) sampled clouds over open ocean near Barrow, AK with the goal to collect observations to advance the understanding of the dynamics and microphysical processes in mixed-phase clouds. This case contains forcing for the entire IOP period.

MPACE Run Script

MPACE-B - Mixed-Phase Stratocumulus Intercomparison

Overview: The Mixed-Phase Arctic Cloud Experiment (MPACE, Verlinde et al. 2007) sampled clouds over open ocean near Barrow, AK with the goal to collect observations to advance the understanding of the dynamics and microphysical processes in mixed-phase clouds. Here we use the same forcing of that described in the intercomparison of Klein et al. (2009), which represents a cold-air outbreak mixed-phase stratocumulus cloud.

Verification Notes: The model simulates solid cloud cover for the duration of the run, with contributions from both cloud liquid and cloud ice

MPACE-B Run Script

RICO - Trade-wind cumulus

Overview: Rain in Cumulus over the Ocean (RICO) was a field campaign that took place during November 2004 through January 2005 in the trades over western Africa (Rauber et al. 2007 ). The main goal of the campaign was to measure processes related to the formation of rain in shallow cumulus and how precipitation modifies the structure and ensembles of trade wind clouds. This case uses the same forcing and setup as that used in the RICO LES intercomparison study of VanZanten et al. (2011) . RICO is a good test of maritime shallow convection, but since its forcing is a composite over many days, it is difficult to compare directly against observations.

Verification Notes: After and initial spin-up period, the model simulates shallow clouds that are representative of a trade-wind cumulus regime. However, the clouds are simulated with an oscillatory behavior which as been identified as a behavior by Xue et al. (2017, submitted) due to the interactions between CLUBB and the microphysics.

RICO Run Script

TOGA-COARE - Tropical Pacific Warm Pool Deep Convection

Overview: This case is based on the tropical ocean - global atmosphere (TOGA) coupled ocean atmosphere response experiment (COARE), which was an IOP that took place from 1 November 1992 through 28 February 1993 in the western pacific warm pool to better understand the principle process responsible for coupling of the ocean and atmosphere. The IOP forcing file for this case begins on 18 December 1992 and is a 22 day case.

TOGA-COARE Run Script

TWP-ICE - Tropical Pacific Warm Pool Deep Convection

Overview: This case is based on the Tropical Pacific Warm Pool International Cloud Experiment (TWP-ICE), which was an intense airborne measurement campaign conducted in the region near Darwin, Northern Australia in early 2006 with the goal to describe the properties of tropical cirrus and the convection that leads to their formation. The IOP forcing file for this case begins on 17 January 2006 and is a 27 day case, and described in Xie et al. (2010).

Verification Notes: This case is simulated with the expected behavior.

TWP-ICE Run Script

Unverified Cases

We have not checked whether the cases below actually run successfully or provide credible output. We are archiving them here in case others need them before we have time to perform any testing.

ARM SGP - Southern Great Plains Site

Overview: This is a series of forcing files from the ARM Southern Great Plains (SGP) site. This is similar to the ARM Continuous Forcing case, but this contains several periods of forcing that falls outside of that window. The script provided for this case is set up to run the period that starts on March, 2000 (forcing file SGP0003_iopfile_4scam.nc).

Note that there are 11 other forcing files available but it is up to the user to enter the correct starting dates and run time in the scripts. This can easily be done by using nco commands with the forcing files. Additional forcing files that are available:

  1. SGP9704_iopfile_4scam.nc (Start date: April 2, 1997; Duration: 21 days)
  2. SGP9706_iopfile_4scam.nc (Start date: June 18, 1997; Duration: 29 days)
  3. SGP9709_iopfile_4scam.nc (Start date: September 15, 1997; Duration: 21 days)
  4. SGP9804_iopfile_4scam.nc (Start date: April 27, 1998; Duration: 21 days)
  5. SGP9901_iopfile_4scam.nc (Start date: January 19, 1999; Duration: 21 days)
  6. SGP9903_iopfile_4scam.nc (Start date: March 1, 1999; Duration: 21 days)
  7. SGP0003_iopfile_4scam.nc (Start date: March 1, 2000; Duration: 21 days)
  8. SGP0009_iopfile_4scam.nc (Start date: September 25, 2000; Duration: 13 days)
  9. SGP0011_iopfile_4scam.nc (Start date: November 27, 2000; Duration: 2 days)
  10. SGP0205_iopfile_4scam.nc (Start date: May 25, 2002; Duration: 21 days)
  11. SGP0211_iopfile_4scam.nc (Start date: November 6, 2002; Duration: 20 days)
  12. SGP0305_iopfile_4scam.nc (Start date: May 1, 2003; Duration: 30 days)

ARM SGP Run Script

ATEX - Trade Wind Cumuli Under a Strong Inversion

Overview: This case is based on the Atlantic Trade Wind Experiment (ATEX) that took place in February 1969 in the Atlantic northeast trade wind region. The forcing for this case follows that of the LES intercomparison study of Stevens et al. (2001) and represents a case of cumulus under broken stratocumulus.

ATEX Run Script

DARWIN - Forcing from tropical western pacific

Overview This is a series of forcing files from the ARM Tropical Western Pacific (TWP) Ocean site, located in Darwin, Australia. Note that the forcing for this site is not continuous and broken into 11 segments. The script provided below is configured to run with the first period of forcings taken from the November 03, 2004 for a five day period.

Should the user want to run a different period it is up to the user to enter the correct starting dates and run time in the scripts. This can easily be done by using nco commands with the forcing files. Additional forcing files that are available:

  1. DARWIN_0405_p1_iopfile_4scam.nc (Start date: November 3, 2004; Duration: 5 days)
  2. DARWIN_0405_p2_iopfile_4scam.nc (Start date: November 11, 2004; Duration: 45 days)
  3. DARWIN_0405_p3_iopfile_4scam.nc (Start date: January 6, 2005; Duration: 4 days)
  4. DARWIN_0405_p4_iopfile_4scam.nc (Start date: January 17, 2005; Duration: 6 days)
  5. DARWIN_0405_p5_iopfile_4scam.nc (Start date: January 27, 2005; Duration: 29 days)
  6. DARWIN_0405_p6_iopfile_4scam.nc (Start date: February 28, 2005; Duration: 36 days)
  7. DARWIN_0506_p1_iopfile_4scam.nc (Start date: November 10, 2004; Duration: 156 days)
  8. DARWIN_0506_p2_iopfile_4scam.nc (Start date: April 18, 2005; Duration: 12 days)
  9. DARWIN_0607_p1_iopfile_4scam.nc (Start date: October 12, 2004; Duration: 39 days)
  10. DARWIN_0607_p2_iopfile_4scam.nc (Start date: November 24, 2004; Duration: 42 days)
  11. DARWIN_0607_p3_iopfile_4scam.nc (Start date: January 8, 2005; Duration: 8 days)

DARWIN Run Script

RACORO - Continental Boundary Layer Clouds

Overview: The routine ARM Aerial Facility (AAF) Clouds with Low Liquid Water Depths (CLOWD) Optical Radiative Observations (RACORO, Vogelmann et al. (2012)) field campaign was conducted in Oklahoma between January and July of 2009 with the goal of conducting long-term routine flights in the boundary layer liquid water clouds over the DOE ARM SGP site. This IOP focuses on a 28 day period in May of 2009, which predominately consisted of cumulus convection in the observations. The forcing was created from the NOAA Rapid Update Cycle (RUC) analyses that were constrained by the ARM surface observations through a variational analysis approach described in Xie et al. (2004) .

RACORO Run Script

SPARTICUS - Cirrus

Overview: This case is based on the Small Particles in Cirrus Clouds (SPARTICUS) field campaign, which took place at the ARM Climate Research Facility from December 2009 through June 2010. The IOP forcing for this case begins on April 1, 2010 and is 30 days in duration. The forcing was created from the NOAA Rapid Update Cycle (RUC) analyses that were constrained by the ARM surface observations through a variational analysis approach described in Xie et al. (2004).

SPARTICUS Run Script

E3SM Replay Mode

REPLAY - Replay a column from an E3SM run

Overview This functionality provides the ability to Replay a column from a full E3SM simulation. This feature is NOT available in the E3SMv1 release, however it is available in the E3SM master as of January 9, 2019. NOTE that currently the E3SM SCM REPLAY option CANNOT provide bit-4-bit reproducibility with an E3SM run (this feature will be available in a future PR). It can, however, reproduce the behavior of an E3SM column, which is appropriate for most scientific applications.

Running the E3SM REPLAY option is a multiple phase process:

  1. You need to generate the forcing from an E3SM simulation. To do this, configure your E3SM simulation with the directive "-camiop". By default, this will generate forcing data at EVERY E3SM time step and for the entire globe and place that data in the h1 history tape. This can add up to lots of data very quickly. Therefore, if there is a particular region of the globe you are interested in, it is strongly advised that you produce forcing for a subset of the globe with the "fincl2lonlat" namelist option. You can find documentation on this namelist variable in the components/cam/bld/namelist_files/namelist_definition.xml file.

NOTE: It is advisable to NOT add any additional data to fincl2 in your namelist and make sure you are writing at the frequency you want write at. Set nhtfrq(2) = 1 to output every E3SM timestep (though you don't have to). Also, it is advised to set mfilt(2) to a large number to ensure all your forcing is in one file (else, they will have to be stitched together manually post-run).

  1. If you produced forcing data for the entire globe then go directly to step 3. If you produced data for a subset of the globe then doing an ncdump -h on your .cam.h1 file you will see that E3SM appended each variable name with the latlon domain bounds, in addition to creating new dimension names. These need to be renamed before the SCM can run these. A script is provided below, which can be used to produce a file that the SCM can you. You must have nco loaded on your machine before you can run this script.

SCM Replay conversion script

  1. Run the E3SM SCM with your forcing file. You should be using the same E3SM code tag in which you generated the forcing file from. Please use the E3SM SCM Replay script below and fill out all the required information in the USER DEFINED settings. After you build the model, you should verify that your namelist settings from your global run and SCM run are as close as possible (compare the atm_in files from the two runs). If there are significant differences, then change your namelist in your SCM REPLAY script and compile and run again.

REPLAY Run Script

TIPS: If this is your first time running the REPLAY option it is STRONGLY advised that you test out a short simulation first. For example, run E3SM for a day or two with your output and make sure this runs in the SCM.

You can’t perform that action at this time.