We study the adjustment of the tropical atmosphere to localized surface heating using a Lagrangian atmospheric model(LAM) that simulates a realistic Madden-Julian Oscillation (MJO) – the dominant, eastward-propagating mode of tropical intraseasonal variability modulating atmospheric convection. Idealized warm sea surface temperature (SST) anomalies of different aspect ratios and magnitudes are imposed in the equatorial Indian Ocean during MJO-neutral conditions and then maintained for 15 days. Throughout these experiments, we observe a robust generation of an MJO event, evident inprecipitation, velocity, temperature and moisture fields, which becomes a key element of atmospheric adjustment along with the expected Kelvin and Rossby waves. The MJO circulation pattern gradually builds up during the first week, and then starts to propagate eastward at a speed of 5-7m/s. The upper-level quadrupole circulation characteristic of the MJObecomes evident around day 14, with two anticyclonic gyres generated by the Gill-type response to convective heating andtwo cyclonic gyres forced by the excited Kelvin waves and extratropical Rossby wave trains. A moisture budget analysis shows that the eastward propagation of the MJO is controlled largely by anomalous advection of moisture and by the residual between anomalous moisture accumulation due to converging winds and precipitation. The initial MJO event is followed bysuccessive secondary events, maintaining the MJO for several more cycles. Thus, for the first time to our knowledge, this study highlights the fundamental role of the MJO played in the adjustment of the moist equatorial atmosphere to heating.  

This dataset contains 6-hourly model output from 1) the Control experiment, which is integrated for two years; 2) 1ºC experiments where the magnitude of the SST perturbation is 1ºC; 3) 2ºC experiments where the magnitude of the SST perturbation is 2ºC and 4) 5ºC experiments where the magnitude of the SST perturbation is 5ºC.

The horizontal resolution is the data is 6º-longitude by 6º-latitude. The vertical resolution is 50hPa. The temporal resolution is 6 hours. 

In 1 (2, 5) ºC experiments, there are 3 sets of experiments, namely Aspect1, Aspect3 and Aspect 5 which indicate the spatial scale of the SST perturbatin. Take 1ºC-Aspect1 experiments for example, it consists of 20 ensemble experiments, each initiated on day 260, 260.5, 261, 261.5 ...... 269.5 of the Control experiment and integrated until day 350.

We include the dynamical (e.g. wind speed, moisture, temperature), precipitation and surface evaporation fields for the Control experiment and 2ºC-Aspect3 experiments, which are the focus of the paper.

We include U200hPa, precipitation fields for the other perturbation experiments. 

%To read dynamical field of the Control experiment between day 260.25 and 350, use Matlab command as below:

a = importdata('Control_dynamic_day260p25_350.dat');

lon = a.lon; % longitude from 3ºE, 9ºE to 357ºE (3ºE), resolution is 6º longitude

lat = a.lat; % latitude from -87ºN (87ºS) to 87ºN, resolution is 6º latitude

p = a.p; % pressure level from 50hPa to 1000hPa, resolution of 50 hPa.

u = a.u; %zonal velocity (m/s) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

v = a.v; % meridional velocity (m/s) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

T = a.T; % temperature (Kelvin) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

q = a.q; % specific humidity (kg/kg) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

 

%To read the precipitation field of the Control experiment in the first year, use Matlab command as below:

a = importdata('Control_prec_year1.dat');

lon = a.lon; % longitude from 3ºE, 9ºE to 357ºE (3ºE), resolution is 6º longitude

lat = a.lat; % latitude from -87ºN (87ºS) to 87ºN, resolution is 6º latitude

prec = a.prec; % integrated precipitation in the past 6 hours (mm). % (longitude,latitude,time), (60, 30, 365*4).

%To get day 1 total precipitation for example 

prec_day1 = sum(prec(:,:,1:4),3);

 

%To read the surface evaporation field of the Control experiment in the first year, use Matlab command as below:

a = importdata('Control_evap_year1.dat');

lon = a.lon; % longitude from 3ºE, 9ºE to 357ºE (3ºE), resolution is 6º longitude

lat = a.lat; % latitude from -87ºN (87ºS) to 87ºN, resolution is 6º latitude

evap = a.evap; % integrated evaporation in the past 6 hours (mm). % (longitude,latitude,time), (60, 30, 365*4).

%To get day 1 total evaporation for example

evap_day1 = sum(evap(:,:,1:4),3);

 

% To get the dynamical field in the 2ºC-Aspect3-ens1 for example,

a = importdata('2K_Aspect3_dynamic_ens1.dat');

lon = a.lon; % longitude from 3ºE, 9ºE to 357ºE (3ºE), resolution is 6º longitude

lat = a.lat; % latitude from -87ºN (87ºS) to 87ºN, resolution is 6º latitude

p = a.p; % pressure level from 50hPa to 1000hPa, resolution of 50 hPa.

u = a.u; %zonal velocity (m/s) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360) %because ens1 is integrated from day 260 to 350, 4*(350 - 260) = 360;

v = a.v; % meridional velocity (m/s) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

T = a.T; % temperature (Kelvin) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

q = a.q; % specific humidity (kg/kg) %dimension (longitude,latitude,pressure,time) (60, 30, 20, 360)

 

%To read the precipitation field of the 2ºC-Aspect3-ens1 experiment, use Matlab command as below:

a = importdata('2K_Aspect3_prec_ens1.dat');

lon = a.lon; % longitude from 3ºE, 9ºE to 357ºE (3ºE), resolution is 6º longitude

lat = a.lat; % latitude from -87ºN (87ºS) to 87ºN, resolution is 6º latitude

prec = a.prec; % integrated precipitation in the past 6 hours (mm). % (longitude,latitude,time), (60, 30, 365*4).

%To get day 1 total precipitation for example 

prec_day1 = sum(prec(:,:,1:4),3);