Global mapping of greenhouse gases retrieved from GOSAT Level 2 products by using a kriging method

Figures & data

Table 1. Major parameters of TANSO-FTS.

Pointing mechanism	Configuration	Two-axis (wholly redundant)
	Scanning	CT (±35°) AT (±20°)
	View	IFOV <10.5 km; scan width 790 km
Interferogram	Scanning frequency	0.25 s^–1, 0.5 s^–1, or 1 s^–1
		SWIR			TIR
	Band	1	2	3	4
	Polarization	Applicable	Applicable	Applicable	N/A
	Wave number (cm^–1)	12,900–13,200	5800–6400	4800–5200	700–1800
	Wavelength (µm)	0.75–0.78	1.56–1.72	1.92–2.08	5.5–14.3
	Resolution (cm^–1)	0.2 cm^–1 (scan on both sides) (MOPD ±2.5 cm)
	Detector	Si	InGaAs		PC-MCT

Figure 1. Nominal footprints of TANSO-FTS in the vicinity of Japan.

Figure 2. Definitions of technical terms used in the method. The circles are semi-variograms and the solid curve is a semi-variogram model obtained by fitting a curve to the circles.

Figure 3. Flowchart of the steps involved in creating FTS Level 3 products. The numbers in the boxes denote the section numbers in the manuscript.

Figure 4. An example of empirical semi-variograms. Open circles and triangles denote semi-variograms obtained from data over ocean and land, respectively. Solid curves show empirical semi-variogram curves obtained by least-squares fitting of each semi-variogram plot.

Figure 5. Global XCO₂ distributions simulated with the NIES.05 atmospheric tracer transport model. The average column density for one month is plotted in each panel. Each panel shows the monthly distribution in one of four representative months (seasons) in a year.

Figure 6. Virtual XCO₂ distributions simulated with the NIES.05 atmospheric tracer transport model.

Figure 7. Semi-variograms and semi-variogram curves estimated from the virtual L2 data shown in Figure 6. Each open circle and triangle was obtained from data over ocean and land, respectively. Solid curves show semi-variogram models obtained by least-squares fitting of each semi-variogram. The values calculated by the models are used to construct semi-variogram matrices. Each panel corresponds to the analogous panel in Figure 5.

Figure 8. L3 maps calculated from the virtual L2 maps shown in Figure 6. Each panel corresponds to the analogous panel in Figure 5.

Figure 9. Standard deviation maps for the L3 maps shown in Figure 6. Each panel corresponds to the analogous panel in Figure 5.

Figure 10. Difference between the ‘ideal’ XCO₂ distribution shown in Figure 5 and the reconstructed XCO₂ distribution obtained by kriging (Figure 8). The white zones are areas where values cannot be determined by the method used there. Each panel corresponds to the analogous panel in Figure 5.

Figure 11. Histograms of the data plotted in Figure 10. Each graph corresponds to the analogous panel in Figure 5.

Figure 12. Real L2 maps retrieved from GOSAT Level 1 B products.

Figure 13. Semi-variograms and semi-variogram curves estimated from the real L2 maps shown in Figure 12.

Figure 14. Real L3 maps estimated from the real L2 maps shown in Figure 12.

Figure 15. Standard deviation maps for the L3 maps shown in Figure 14.

Figure 16. Estimation error, , for the real data.