Influence of meteorological parameters on atmospheric CO₂ at Bharati, the Indian Antarctic research station

Figures & data

Figure 1. The location of Bharati, the Indian research station, in Antarctica.

Table 1. Details of the data used and their sources. The period was 22 January–25 February 2016.

Parameter	Data collection frequency	Source
CO2	1 sec	Li-Cor CO2/H2O analyser
AT, RH, WS, wind direction and surface pressure	1 min	Automatic weather station: AT & RH (Rotronic); WS & wind direction (Gill Instruments) and surface pressure sensor (Thies Clima), India Meteorological Department
Precipitation	Daily	Snow gauge, India Meteorological Department
PAR and BLH	Daily	European Centre for Medium-Range Weather Forecasts (http://data-portal.ecmwf.int)
Backward trajectory	Averaged summer season	HYSPLIT (https://ready.arl.noaa.gov/HYSPLIT.php)

Figure 2. Daily variation of atmospheric CO₂ with meteorological parameters (a) AT (°C), (b) RH (%), (c) WS (m s⁻¹) and (d) surface pressure (hPa).

Figure 3. Box and whisker plots of atmospheric CO₂ with meteorological parameters (a) AT (°C), (b) RH (%), (c) WS (m s⁻¹) and (d) wind direction (deg.). The lower and upper whiskers shows 5th and 95th percentiles of the data. The lower and upper quartiles of the vertical boxes represent the 25th and 75th percentiles, respectively – together they comprise the middle 50% of the data. The horizontal line in each vertical box indicates the median of the data. Values which are beyond whiskers are outliers. Lowest and highest CO₂ values are represented at the 5th and 95th percentiles of the data. (e) Frequency distribution of WS and mean CO₂. (f) Precipitation influence on daily mean CO₂.

Table 2. Statistical correlation between CO₂ and its influencing parameters at Bharati Station.

		PAR (MJ m^−2)	Surface pressure (hPa)	AT (°C)	BLH (m)	WS (m s^−1)	RH (%)	Precipitation (mm)
CO2 (ppm)	r	0.210	0.230	0.130	−0.110	−0.180	−0.320	−0.550
	p level	0.242	0.205	0.468	0.512	0.303	0.071	0.003
	n	33	33	33	33	33	33	27

Figure 4. Correlation of CO₂ with different meteorological parameters at Bharati station, during summer 2016.

Table 3. Training and results of the least squares multiple regression analysis performed between meteorological parameters and the CO₂ mixing ratio (the dependent variable).

Selection method		Forward: significant variables selected sequentially
Significance test		Reject if p > 0.10		Accept if p < 0.05
Sample size (n)		27
Coefficient of determination (R^2)		0.29
Multiple correlation coefficient (r)		0.55
Results of regression equation	Independent variables	Coefficient	Constant	Standard error	p value
	Precipitation	−0.75	396.05	0.23	0.003
	Variables rejected in the significance test	PAR, pressure, AT, BLH, WS and RH

Figure 5. Long range air-mass trajectories to Bharati at altitudes (a) 3 km, (b) 2 km, (c) 1 km altitude and (d) change in CO₂ (%) in different wind sectors.