TEMPO NO 2 Validation Ron Cohen UC Berkeley
TEMPO NO 2 Validation Ron Cohen, UC Berkeley
Contract requirement 1. Precision of 1 x 1015 molecules/cm 2 (~0. 5 ppb in the PBL) Approach: ~3 Pandoras for 1 month; 4 seasons
NO 2 Validation issues Most approaches to using the data assume/will work better if the observations have little bias (or a Gaussian distribution of bias). We want the data to be unbiased with respect to viewing and solar zenith angles (time of day), cloudiness, aerosol, albedo (several comments about this yesterday).
Los Angeles: WRF-Chem
Particulate Matter (co-emitted with CO 2, NOx, CO, …) observations modeled fit 1σ variation range from Choi et al. 2014
http: //behr. cchem. berkeley. edu// NASA standard Terrain pressure BEHR High-res terrain database, center of OMI footprint High-res terrain database, average over OMI footprint Terrain reflectivity Monthly 1° × 1° MODIS, 8 day 0. 05° × 0. 05° NO 2 profile shape Annually 2° × 2. 5° WRF-Chem, Monthly 4 × 4 km 2 (CA&NV) 12 x 12 km 2 U. S. Clouds OMI cloud product MODIS cloud product Russell et al. , Atmos Chem & Phys 11, 8543 -8554, 2011
Terrain Reflectivity (Albedo) MODIS True Color SP NO 2 June 18, 2008 OMI Monthly Albedo NASA Standard Product June 2008 MODIS 8 day Albedo BEHR June 2008 Russell et al. , Atmos Chem & Phys, 2011
Terrain Reflectivity (Albedo) Histogram of systematic errors Russell et al. , Atmos Chem & Phys, 2011
NO 2 profile shape Histogram of systematic errors Russell et al. , Atmos Chem & Phys, 2011
The BEHR product is generally higher in urban regions and lower in rural regions than the operational products Standard Product BEHR % Difference Russell et al. , Atmos Chem & Phys, 2011
Trends in cities are similar while trends at power plants are more variable 47 cities, 23 power plants! Russell et al. , ACP 2012
NO 2 Validation Strategies Check all possible avenues for internal consistency Example: look in remote places with uniform (but low) NO 2 columns and make sure observed variation is geophysical sensible—not driven by viewing angle etc. Stare at one location for an hour (at midday) and check that clouds moving across the scene don’t affect the interpretation. Examine repeats at a power plant with near constant emissions and check that there is little variation of NO 2 with time of day.
May–October 2005– 2006 0 1 2 3 4 5 6 7 8 NO 2 (molecules cm– 2) 9 10 x 1015 OMI Berkeley High-resolution Retrieval (BEHR)
NO 2 Validation Strategies Additional “conventional data” Aircraft/ground based experiments e. g. DISCOVER; KORUS Surface network additional PANDORA’s
NO 2 Validation Strategies “unconventional data”
CO 2 Emissions in San Francisco bay area at 1 km resolution
NO NO 2 O 3 CO CO 2 aerosol
BEACO 2 N observing network http: //beacon. berkeley. edu/
Electrochemical O 3, NO 2 & CO Sensors Shinyei Grove Particulate Sensor Vaisala GMP 343 NDIR CO 2 Sensor
BEACO 2 N CO 2 2013 Sites: Burckhalter Prescott Laurel Kaiser College. Prep Korematsu ODowd St. Liz Head. Royce El. Cerrito NOakland
WRF-STILT for day bridge was closed 10 km Alex Turner
NO 2 Validation Strategies “other unconventional data? ” Profiling with small sensors and drones LIDARS Sondes
- Slides: 22