Tropospheric Emissions: Monitoring of Pollution (TEMPO)

Abstract

TEMPO was selected in 2012 by NASA as the first Earth Venture Instrument, for launch between 2018 and 2021. It will measure atmospheric pollution for greater North America from space using ultraviolet and visible spectroscopy. TEMPO observes from Mexico City, Cuba, and the Bahamas to the Canadian oil sands, and from the Atlantic to the Pacific, hourly and at high spatial resolution (~2.1 km N/S×4.4 km E/W at 36.5°N, 100°W). TEMPO provides a tropospheric measurement suite that includes the key elements of tropospheric air pollution chemistry, as well as contributing to carbon cycle knowledge. Measurements are made hourly from geostationary (GEO) orbit, to capture the high variability present in the diurnal cycle of emissions and chemistry that are unobservable from current low-Earth orbit (LEO) satellites that measure once per day. The small product spatial footprint resolves pollution sources at sub-urban scale. Together, this temporal and spatial resolution improves emission inventories, monitors population exposure, and enables effective emission-control strategies. TEMPO takes advantage of a commercial GEO host spacecraft to provide a modest cost mission that measures the spectra required to retrieve ozone (O₃), nitrogen dioxide (NO₂), sulfur dioxide (SO₂), formaldehyde (H₂CO), glyoxal (C₂H₂O₂), bromine monoxide (BrO), IO (iodine monoxide),water vapor, aerosols, cloud parameters, ultraviolet radiation, and foliage properties. TEMPO thus measures the major elements, directly or by proxy, in the tropospheric O₃ chemistry cycle. Multi-spectral observations provide sensitivity to O₃ in the lowermost troposphere, substantially reducing uncertainty in air quality predictions. TEMPO quantifies and tracks the evolution of aerosol loading. It provides these near-real-time air quality products that will be made publicly available. TEMPO will launch at a prime time to be the North American component of the global geostationary constellation of pollution monitoring together with the European Sentinel-4 (S4) and Korean Geostationary Environment Monitoring Spectrometer (GEMS) instruments.

Figure 1:

Average tropospheric column NO2 for 2005–2008 measured from the OMI satellite over the TEMPO field of regard.

Figure 2:

TEMPO footprints overlaid on the Baltimore-Washington metropolitan area. The footprint size here is approximately 2.5 km N/S × 5 km E/W. Map created using Google Earth/Landsat Imagery.

Figure 3.

TEMPO instrument functional block diagram. FPGAs – Field-Programmable Gate Arrays. DITCE – Differential Impedance Transducer Conditioning Electronics.

Figure 4.

Optical ray trace for the TEMPO instrument, including telescope and spectrometer.

Figure 5.

The spectral regions to be measured by TEMPO are illustrated with reflectance spectra for the range of surface and atmosphere scenes using reflectances derives from European Space Agency GOME-1 measurements. The dashed blue boxes indicate the TEMPO spectral coverage.

Figure 6.

The TEMPO INR solution creates a Level-1 (L1) product with geographic metadata for each pixel using smoothed Kalman Filter states. Scan tailoring coefficients compensate for deterministic pointing errors to assure efficient coverage of Greater North America.

Figure 7.

Nominal daily operations for TEMPO instrument.

Figure 8.

Rows of two averaging kernel matrices based on iterative nonlinear retrievals from synthetic TEMPO radiances with the signal to noise ratio (SNR) estimated using the TEMPO SNR model at instrument critical design review in June 2015 for (a) UV (290–345 nm) retrievals and (b) UV/Visible (290–345 nm, 540–650 nm) retrievals for clear-sky condition and vegetation surface with solar zenith angle 25°, viewing zenith angle 45° and relative azimuthal angle 86°. DFS is degrees of freedom for signal, the trace of the averaging kernel matrix, which is an indicator of the number of pieces of independent information in the solution.

Figure 9.

Longitude-altitude cross-section of ozone concentrations (36°N, 3 GMT on June 14, 2010) associated with a stratospheric intrusion. The “true” state from an independent chemical transport model (top left) is compared to the GEOS-Chem model without data assimilation (top right) and with assimilation of TEMPO data (bottom). Local topography is shown in white. Modified from Zoogman et al. [2014].

Figure 10.

SNRs of various artificial lighting types normalized to a common VIIRS-DNB response with a 10s dwell, no co-addition, and no spatial binning, assuming a dark current of 2800 e⁻ s⁻¹. City lights over all of greater North America can either be observed piecemeal over several days or in a single scan near the winter solstice.