Satellite NO₂ as an Economic Indicator for Myanmar

NO₂ Trend (OMI, 2012–2025)

The OMI instrument provides a 13-year monthly record of tropospheric NO₂ over Myanmar. There’s a noticeable decline in NO₂ levels starting in early 2021, following the coup.

The figure below shows the OMI NO₂ time series decomposed into trend, seasonal, and residual components using STL decomposition. The trend component confirms a structural decline in NO₂ beginning around the coup. Seasonal pattern is roughly constant across the 13-year period, consistent with metereological drivers of NO₂.

Seasonal Pattern

NO₂ in Myanmar exhibits strong seasonality, where concentrations peak in the dry season (Oct–Mar) when precipitation-driven washout is minimal, while the monsoon (Jun–Sep) suppresses NO₂ levels through wet deposition.

GDP vs NO₂ Trends by Sector

The following figures show the standardized (z-scored) quarterly OMI NO₂ alongside GDP across all sectors. NO₂ and GDP move broadly together for Manufacturing and Construction, which are the two most combustion-intensive sectors. Mining and Trade show weaker visual co-movement but similar seasonal patterns.

Show code cell outputs Hide code cell outputs

The following figures show quarter-over-quarter and year-over-year percentage changes in OMI NO₂ and GDP. The co-movement is less visually apparent in the growth rates, but there are still some periods of alignment.

NO₂–GDP Regressions

We estimate sector-by-sector log-log OLS regressions, first without and then with quarter dummies.

Model 1: gdp_log ~ log_no2 (all sectors including Energy)
  Communication: N=53
  Construction: N=53
  Electricity: N=53
  Energy: N=53
  Financial Insitutions: N=53
  Manufacturing: N=53
  Mining: N=53
  Rental and other services: N=53
  Social and Administrative: N=53
  Trade: N=53
  Transportation: N=53

	Dependent variable: gdp_log
	Communication	Construction	Electricity	Energy	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)	(11)
Intercept	13.458***	17.868***	11.557***	6.930	13.850***	21.838***	13.474***	13.538***	13.277***	15.539***	14.527***
	(1.986)	(2.144)	(1.200)	(7.538)	(1.718)	(2.167)	(2.536)	(1.568)	(1.586)	(2.340)	(1.270)
log_no2	0.018	0.358*	-0.074	-0.493	0.239	0.593***	0.096	0.026	-0.011	0.055	-0.007
	(0.176)	(0.190)	(0.106)	(0.666)	(0.152)	(0.192)	(0.224)	(0.139)	(0.140)	(0.207)	(0.112)
Observations	53	53	53	53	53	53	53	53	53	53	53
R2	0.000	0.065	0.009	0.011	0.046	0.158	0.004	0.001	0.000	0.001	0.000
Adjusted R2	-0.019	0.047	-0.010	-0.009	0.028	0.142	-0.016	-0.019	-0.019	-0.018	-0.020
Residual Std. Error	0.448 (df=51)	0.483 (df=51)	0.270 (df=51)	1.698 (df=51)	0.387 (df=51)	0.488 (df=51)	0.571 (df=51)	0.353 (df=51)	0.357 (df=51)	0.527 (df=51)	0.286 (df=51)
F Statistic	0.010 (df=1; 51)	3.561* (df=1; 51)	0.489 (df=1; 51)	0.548 (df=1; 51)	2.477 (df=1; 51)	9.578*** (df=1; 51)	0.183 (df=1; 51)	0.035 (df=1; 51)	0.006 (df=1; 51)	0.071 (df=1; 51)	0.004 (df=1; 51)
Note:	*p<0.1; **p<0.05; ***p<0.01

Model 2: 3-Quarter Lag

Model 2: gdp_log ~ log_no2_lag3 (excluding Energy)
  Communication: N=50
  Construction: N=50
  Electricity: N=50
  Financial Insitutions: N=50
  Manufacturing: N=50
  Mining: N=50
  Rental and other services: N=50
  Social and Administrative: N=50
  Trade: N=50
  Transportation: N=50

	Dependent variable: gdp_log
	Communication	Construction	Electricity	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Intercept	11.857***	24.371***	11.205***	12.483***	27.998***	11.748***	14.573***	14.371***	30.221***	21.465***
	(1.635)	(1.632)	(1.145)	(1.601)	(1.548)	(2.478)	(1.468)	(1.460)	(1.010)	(0.851)
log_no2_lag3	-0.129	0.930***	-0.108	0.114	1.135***	-0.061	0.114	0.082	1.351***	0.605***
	(0.145)	(0.144)	(0.101)	(0.142)	(0.137)	(0.219)	(0.130)	(0.129)	(0.089)	(0.075)
Observations	50	50	50	50	50	50	50	50	50	50
R2	0.016	0.464	0.023	0.013	0.589	0.002	0.016	0.008	0.827	0.574
Adjusted R2	-0.004	0.453	0.003	-0.007	0.581	-0.019	-0.005	-0.012	0.823	0.565
Residual Std. Error	0.356 (df=48)	0.356 (df=48)	0.250 (df=48)	0.349 (df=48)	0.337 (df=48)	0.540 (df=48)	0.320 (df=48)	0.318 (df=48)	0.220 (df=48)	0.185 (df=48)
F Statistic	0.796 (df=1; 48)	41.540*** (df=1; 48)	1.133 (df=1; 48)	0.650 (df=1; 48)	68.851*** (df=1; 48)	0.079 (df=1; 48)	0.771 (df=1; 48)	0.405 (df=1; 48)	228.883*** (df=1; 48)	64.656*** (df=1; 48)
Note:	*p<0.1; **p<0.05; ***p<0.01

Model 3: NO₂ + Overall NTL

Model 3: gdp_log ~ log_no2 + ntl_sum_log (excluding Energy, no seasonal controls)
  Communication: N=53
  Construction: N=53
  Electricity: N=53
  Financial Insitutions: N=53
  Manufacturing: N=53
  Mining: N=53
  Rental and other services: N=53
  Social and Administrative: N=53
  Trade: N=53
  Transportation: N=53

	Dependent variable: gdp_log
	Communication	Construction	Electricity	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Intercept	8.158**	21.960***	9.787***	10.729***	28.740***	-1.963	6.409**	6.095**	26.820***	18.207***
	(3.371)	(3.703)	(2.088)	(2.973)	(3.619)	(3.573)	(2.466)	(2.498)	(3.623)	(2.140)
log_no2	-0.000	0.372*	-0.080	0.228	0.617***	0.043	0.001	-0.036	0.094	0.005
	(0.171)	(0.188)	(0.106)	(0.151)	(0.184)	(0.182)	(0.125)	(0.127)	(0.184)	(0.109)
ntl_sum_log	0.379*	-0.292	0.126	0.223	-0.493**	1.103***	0.509***	0.513***	-0.806***	-0.263**
	(0.197)	(0.217)	(0.122)	(0.174)	(0.212)	(0.209)	(0.144)	(0.146)	(0.212)	(0.125)
Observations	53	53	53	53	53	53	53	53	53	53
R2	0.069	0.098	0.030	0.077	0.241	0.360	0.200	0.198	0.225	0.081
Adjusted R2	0.032	0.062	-0.009	0.040	0.210	0.334	0.168	0.166	0.195	0.044
Residual Std. Error	0.436 (df=50)	0.479 (df=50)	0.270 (df=50)	0.385 (df=50)	0.468 (df=50)	0.462 (df=50)	0.319 (df=50)	0.323 (df=50)	0.469 (df=50)	0.277 (df=50)
F Statistic	1.851 (df=2; 50)	2.720* (df=2; 50)	0.781 (df=2; 50)	2.076 (df=2; 50)	7.918*** (df=2; 50)	14.068*** (df=2; 50)	6.256*** (df=2; 50)	6.172*** (df=2; 50)	7.278*** (df=2; 50)	2.209 (df=2; 50)
Note:	*p<0.1; **p<0.05; ***p<0.01

Model 4: Seasonal Controls

Model 4: gdp_log ~ log_no2 + quarter (excluding Energy)
  Communication: N=53
  Construction: N=53
  Electricity: N=53
  Financial Insitutions: N=53
  Manufacturing: N=53
  Mining: N=53
  Rental and other services: N=53
  Social and Administrative: N=53
  Trade: N=53
  Transportation: N=53

	Dependent variable: gdp_log
	Communication	Construction	Electricity	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Intercept	-2.814	12.710***	18.968***	8.487	17.609***	-8.537	3.296	7.771	22.979***	14.954***
	(6.455)	(3.955)	(3.938)	(5.378)	(2.872)	(8.117)	(5.055)	(5.353)	(3.018)	(2.528)
log_no2	-1.496**	-0.137	0.616*	-0.269	0.188	-1.953**	-0.928*	-0.525	0.740**	0.028
	(0.598)	(0.366)	(0.365)	(0.498)	(0.266)	(0.752)	(0.468)	(0.496)	(0.279)	(0.234)
quarter[T.2]	-0.833**	-1.054***	0.462**	-0.650**	-0.963***	-1.215***	-0.625**	-0.387	-0.279*	-0.311**
	(0.344)	(0.211)	(0.210)	(0.286)	(0.153)	(0.432)	(0.269)	(0.285)	(0.161)	(0.135)
quarter[T.3]	-1.468**	-0.647*	0.700**	-0.483	-0.639**	-2.023***	-0.987**	-0.538	0.349	-0.114
	(0.559)	(0.343)	(0.341)	(0.466)	(0.249)	(0.703)	(0.438)	(0.464)	(0.262)	(0.219)
quarter[T.4]	-1.140**	-0.051	0.459*	-0.422	0.188	-1.471**	-0.591*	-0.307	1.160***	0.330*
	(0.445)	(0.272)	(0.271)	(0.371)	(0.198)	(0.559)	(0.348)	(0.369)	(0.208)	(0.174)
Observations	53	53	53	53	53	53	53	53	53	53
R2	0.127	0.737	0.118	0.228	0.878	0.156	0.141	0.059	0.863	0.672
Adjusted R2	0.054	0.715	0.045	0.163	0.868	0.086	0.070	-0.020	0.851	0.645
Residual Std. Error	0.431 (df=48)	0.264 (df=48)	0.263 (df=48)	0.359 (df=48)	0.192 (df=48)	0.542 (df=48)	0.338 (df=48)	0.357 (df=48)	0.202 (df=48)	0.169 (df=48)
F Statistic	1.745 (df=4; 48)	33.632*** (df=4; 48)	1.606 (df=4; 48)	3.535** (df=4; 48)	86.146*** (df=4; 48)	2.216* (df=4; 48)	1.974 (df=4; 48)	0.746 (df=4; 48)	75.365*** (df=4; 48)	24.608*** (df=4; 48)
Note:	*p<0.1; **p<0.05; ***p<0.01

Model 4a: Rainfall Control (No Quarter FE)

We control for the physical mechanism behind NO₂ seasonality by adding rainfall as a control variable instead of quarter fixed effects.

Model 5: QoQ % Change (No Lag)

Model 5: gdp_qoq ~ no2_qoq (excluding Energy)
  Communication: N=52
  Construction: N=52
  Electricity: N=52
  Financial Insitutions: N=52
  Manufacturing: N=52
  Mining: N=52
  Rental and other services: N=52
  Social and Administrative: N=52
  Trade: N=52
  Transportation: N=52

	Dependent variable: gdp_qoq
	Communication	Construction	Electricity	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Intercept	0.140	0.184**	0.019	0.050	0.224**	0.032	0.035	0.026	0.289**	0.091
	(0.092)	(0.085)	(0.020)	(0.045)	(0.110)	(0.027)	(0.024)	(0.021)	(0.127)	(0.064)
no2_qoq	0.077	0.120	-0.062*	0.262***	0.255	0.219***	0.046	0.012	-0.118	-0.078
	(0.157)	(0.147)	(0.034)	(0.077)	(0.189)	(0.046)	(0.042)	(0.036)	(0.219)	(0.110)
Observations	52	52	52	52	52	52	52	52	52	52
R2	0.005	0.013	0.062	0.187	0.035	0.308	0.024	0.002	0.006	0.010
Adjusted R2	-0.015	-0.007	0.043	0.171	0.016	0.294	0.005	-0.018	-0.014	-0.010
Residual Std. Error	0.645 (df=50)	0.600 (df=50)	0.140 (df=50)	0.317 (df=50)	0.772 (df=50)	0.190 (df=50)	0.170 (df=50)	0.149 (df=50)	0.896 (df=50)	0.450 (df=50)
F Statistic	0.239 (df=1; 50)	0.667 (df=1; 50)	3.302* (df=1; 50)	11.519*** (df=1; 50)	1.829 (df=1; 50)	22.278*** (df=1; 50)	1.232 (df=1; 50)	0.118 (df=1; 50)	0.292 (df=1; 50)	0.499 (df=1; 50)
Note:	*p<0.1; **p<0.05; ***p<0.01

Model 6: QoQ % Change (3-Quarter Lag)

Model 6: gdp_qoq ~ no2_qoq_lag3 (excluding Energy)
  Communication: N=49
  Construction: N=49
  Electricity: N=49
  Financial Insitutions: N=49
  Manufacturing: N=49
  Mining: N=49
  Rental and other services: N=49
  Social and Administrative: N=49
  Trade: N=49
  Transportation: N=49

	Dependent variable: gdp_qoq
	Communication	Construction	Electricity	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Intercept	0.126	0.114*	0.026	0.088*	0.123**	0.054	0.018	0.014	0.100**	0.011
	(0.093)	(0.060)	(0.019)	(0.050)	(0.058)	(0.032)	(0.020)	(0.017)	(0.048)	(0.042)
no2_qoq_lag3	-0.019	0.760***	-0.120***	0.020	1.160***	0.078	0.194***	0.122***	1.439***	0.617***
	(0.159)	(0.103)	(0.032)	(0.085)	(0.099)	(0.055)	(0.034)	(0.029)	(0.081)	(0.072)
Observations	49	49	49	49	49	49	49	49	49	49
R2	0.000	0.537	0.234	0.001	0.745	0.041	0.412	0.274	0.870	0.610
Adjusted R2	-0.021	0.527	0.218	-0.020	0.740	0.021	0.400	0.258	0.867	0.601
Residual Std. Error	0.638 (df=47)	0.415 (df=47)	0.128 (df=47)	0.343 (df=47)	0.398 (df=47)	0.222 (df=47)	0.136 (df=47)	0.117 (df=47)	0.327 (df=47)	0.290 (df=47)
F Statistic	0.015 (df=1; 47)	54.412*** (df=1; 47)	14.347*** (df=1; 47)	0.056 (df=1; 47)	137.561*** (df=1; 47)	2.020 (df=1; 47)	32.936*** (df=1; 47)	17.721*** (df=1; 47)	314.456*** (df=1; 47)	73.445*** (df=1; 47)
Note:	*p<0.1; **p<0.05; ***p<0.01

Model 7: YoY % Change (No Lag)

Model 7: gdp_yoy ~ no2_yoy (excluding Energy)
  Communication: N=49
  Construction: N=49
  Electricity: N=49
  Financial Insitutions: N=49
  Manufacturing: N=49
  Mining: N=49
  Rental and other services: N=49
  Social and Administrative: N=49
  Trade: N=49
  Transportation: N=49

	Dependent variable: gdp_yoy
	Communication	Construction	Electricity	Financial Insitutions	Manufacturing	Mining	Rental and other services	Social and Administrative	Trade	Transportation
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)	(9)	(10)
Intercept	0.266**	0.076**	0.034*	0.114***	0.034*	0.185***	0.099***	0.104***	0.008	0.032*
	(0.117)	(0.029)	(0.019)	(0.033)	(0.017)	(0.032)	(0.021)	(0.026)	(0.018)	(0.017)
no2_yoy	-1.714*	-0.454*	0.037	-0.249	-0.105	-0.727***	-0.191	-0.190	0.045	-0.216
	(0.924)	(0.226)	(0.152)	(0.258)	(0.137)	(0.249)	(0.165)	(0.206)	(0.144)	(0.136)
Observations	49	49	49	49	49	49	49	49	49	49
R2	0.068	0.079	0.001	0.020	0.012	0.153	0.028	0.018	0.002	0.051
Adjusted R2	0.048	0.059	-0.020	-0.001	-0.009	0.135	0.007	-0.003	-0.019	0.030
Residual Std. Error	0.818 (df=47)	0.200 (df=47)	0.135 (df=47)	0.228 (df=47)	0.122 (df=47)	0.220 (df=47)	0.146 (df=47)	0.182 (df=47)	0.128 (df=47)	0.120 (df=47)
F Statistic	3.441* (df=1; 47)	4.036* (df=1; 47)	0.061 (df=1; 47)	0.936 (df=1; 47)	0.582 (df=1; 47)	8.523*** (df=1; 47)	1.336 (df=1; 47)	0.847 (df=1; 47)	0.095 (df=1; 47)	2.509 (df=1; 47)
Note:	*p<0.1; **p<0.05; ***p<0.01

Cross-Correlation Analysis

Cross-correlation functions (CCF) between NO₂ and GDP at various lags, to identify the lead/lag structure of the relationship.

Methodological caveat: These are associations, not causal estimates. Omitted variables may confound the relationship. The small sample (~48 quarters per sector) limits statistical power and prevents adding many controls simultaneously.