Method: Differential Absorption Lidar
The challenge that faced the first developpers of DIAL system for DIAL CO2 mixing ratio measurements is the very stringent requirements on measurement accuracy below 1 % in CO2 volume mixing ratio. This corresponds to a precision of about one order of magnitude higher compared to existing active systems for measuring atmospheric constituents, such as water vapor, ozone or methane [Weitkamp, 2005].
DIAL CO2 mixing ratio accuracy and precision depends on instrumental (for Lidar CO2 differential absorption measurement), spectroscopic and meteorological data accuracies and precisions.
Bruneau et al. (2006) have shown an optimal optical depth of ~ 1.3 for Heterodyne DIAL measurements. CO2 absorption lines in the 2 µm domain enable to respect this condition in the boundary layer (z < 2 km). The CO2 differential absorption measurement is provided by mean on- and off-line return power in a range gate ΔR at the wavelength i, over Mp shots averaged.
α can calculated using a linear fit weighted by standard deviation on the optical depth called the « slope method ». This represents the maximum likelihood estimate of the mean CO2 differential absorption in the DIAL path (Gibert et al. 2006) (Fig. 2). WF is calculated using new spectroscopic data from diode laser spectroscopy (Joly et al., 2007) which enables to get 0.5 % of accuracy on CO2 differential cross-section. Meteorological data p, T, ρW (water vapor mixing ratio) are provided by ground-based in-situ sensors for horizontal DiAL measurements and MM5 mesoscale model outputs for vertical DiAL measurements.
Fig. 2: Optical depth as a function of range (from Gibert et al. 2006). Slope method.
First DIAL CO2 mixing ratio measurements
First DIAL CO2 mixing ratio measurements were made horizontally in 2004 (Koch et al. 2004; Gibert et al., 2006). In September 2006, we made a field experiment CIEL Comparison In-situ sEnsors and Lidar for co2 mixing ratio measurements to validate DIAL measurements (Gibert et al. 2008a, 2008b). The 2-µm coherent DIAL system (EMIL-LIDIA)provides aerosol backscatter signal, radial velocity and mean CO2 mixing ratio over a 1-km horizontal long path (red line). DIAL measurements were compared with different in-situ sensors: a laser diode spectrometer from GSMA group at Reims University (blue line), a LICOR NDIR analyzer (green line) and flasks (green asteriks) from RAMSES observatory. The yellow line corresponds to gas chromater routine measurements at LSCE laboratory, 5 km away. The standard deviation and bias of mean 1-km CO2 DIAL measurements (compared with the LICOR) are 3.3 ppm and -0.54 ppm for 30 min of time averaging.
Fig. 3: Experimental site during the CIEL field campain and comparison of CO2 mixing ratio measurements from EMIL-LIDIA (red line) and in-situ sensors. From Gibert et al. (2008b). Time and space resolution of DIAL CO2 mean mixing ratio measurements are respectively 30 min and 1 km. Blurred area corresponds to the standard deviation of each instrument.
During the day time and statically unstable condition, the boundary layer is well mixed over 10 km and DIAL and in-situ sensors show similar CO2 mixing ratios. During the night time, the flow becomes statically stable and small-scale vertical (over 0.1 km) and horizontal (over 1 km) gradients are observed. the results illustrate the complexity of inferring surface fluxes of CO2 from atmospheric budgets in the stable suburban boundary layer.
Vertical DIAL measurements
Vertical measurements were achieved in June 2005 (Fig. 4). Before 0930 UTC, DIAL CO2 measurements (red line) are made in the residual layer whereas the ground-based in-situ sensor provides a CO2 mixing ratio in the nocturnal layer. After 0930 UTC, vertical velocities show that both measurements are made in the mixed layer. Within DIAL standard deviation, the two instruments are in good agreement. We show in Table 1 a list of statistical and systematic errors for the DIAL instrument (Gibert et al., 2008).
Fig. 4: Coherent DIAL measurement on June, 10, 2005. (a) Off-line backscatter signal (Log(P.z2)) (b) vertical velocity. Range and time resolution are 75 m and 2 min (c) DIAL CO2 mixing ratio measurements compared with the in-situ (GC) routine measurements at RAMSES observatory. Time and space resolution are 30 min and 1 km.
Table 1: Statistic and systematic uncertainties of vertical coherent DIAL mean CO2 mixing ratio measurements in the ABL
Free troposphere measurements. IPDA technique using clouds reflectivity
Using the return power from cloud reflectivity, we also made range resolved measurements in the boundary layer and in the free troposphere using both the slope method and the IPDA technique. Knowing the atmospheric boundary layer (ABL) CO2 mixing ratio, the optical depth in the cloud and the integral of ABL and free troposphere weighting functions (SWF), we calculated the free tropospheric CO2 mixing ratio:
DIAL free tropospheric CO2 mixing ratio agrees well with airborne in-situ sensor despite a large standard deviation (7 ppm) for standard CO2 troposhere variability (Fig. 5).
Fig. 5: Coherent DIAL measurement on November, 5, 2004. (a) Off-line backscatter signal (Log(P.z2)) (b) DIAL CO2 mixing ratio measurements compared with the in-situ (GC) routine measurements at RAMSES observatory. Time and space resolution are 30 min and 1 km for ABL DIAL measurements. Mean CO2 mixing ratio in the free troposphere are indicated from the IPDA technique (*) and for airborne in-situ sensors measurements (dotted line). From Gibert et al. 2008