OMI validation status

18
OMI-Science Team Meeting | 11 March 2014 OMI validation status Ankie Piters, KNMI

description

OMI validation status. Ankie Piters, KNMI. What is the status of OMI validation?. H ow good is our knowledge about the accuracy/quality of the data? (rather than: “how good is our data”) How well is our knowledge communicated to the user? A product is “validated” when: - PowerPoint PPT Presentation

Transcript of OMI validation status

Page 1: OMI validation status

OMI-Science Team Meeting | 11 March 2014

OMI validation status

Ankie Piters, KNMI

Page 2: OMI validation status

OMI-Science Team Meeting | 11 March 20142

What is the status of OMI validation?

How good is our knowledge about the accuracy/quality of the data? (rather than: “how good is our data”)How well is our knowledge communicated to the user?

A product is “validated” when:

The newest version of a product has been compared on a global scale to different sources of known accuracy, and for all seasons. Possible systematic effects have been studied with respect to instrumental, algorithmic, or atmospheric parameters. The results of these studies are quantified in terms of e.g. a bias and precision, and their dependence on certain parameters. And regimes with different error-characteristics are specified.The results are published in peer-reviewed literature, and summarised/referred to in the product README files.

Page 3: OMI validation status

Questions to Product Developers:• To what correlative data has your product been compared?• Has your product been compared to data at different latitude bands?• Has your product been compared for different time periods or seasons?• Is it possible to state an average bias, precision, or accuracy for your product?• Do the validation studies show systematic effects or dependencies w.r.t. certain regions, atmospheric circumstances, or instrumental parameters?• Can you specify certain regimes where your product is less accurate, or should be used with caution?• What knowledge is missing about the quality, accuracy of your product?

OMI-Science Team Meeting | 11 March 20143

Page 4: OMI validation status

4

bias / precision

green: quantified

orange: not yet

yellow: missing info, or waiting for validation of current versioncorellative data

green: more independent validation sources

orange: less validation sources

version

green: Latest validated version

orange: Newer version than the one validated

seasonal comparisons

green: yes

orange: not all seasons covered

additional information

blue: known issues

yellow: missing info

(*) probably solved in next version

results published

green: peer-reviewed paper(s)

orange: other

global comparisons

green: yes

orange: large latitude range not covered

Page 5: OMI validation status

OMI-Science Team Meeting | 11 March 2014 5

OMDOAO3 - GlobalBrewer Dobson

bias w.r.t Brewer: -1/-2%; precision: 3%

comp to Dobson indicate a weak slope as a function of latitude

latitude latitude

Ozone columns - KNMI

source: http://lap.physics.auth.gr/eumetsat/totalozone

Page 6: OMI validation status

OMI-Science Team Meeting | 11 March 2014 6

OMDOAO3 - SZABrewer Dobson

negative bias compared to Brewers increase at SZA>75o, to -3% at 85o

sza szasource: http://lap.physics.auth.gr/eumetsat/totalozone

Ozone columns - KNMI

Page 7: OMI validation status

OMI-Science Team Meeting | 11 March 2014 7

OMDOAO3 – Timeseries NHBrewer Dobson

small trend over time of +0.1% per year

source: http://lap.physics.auth.gr/eumetsat/totalozone

Ozone columns - KNMI

Page 8: OMI validation status

Zonal mean stratospheric NO2

Northern hemis-phere (35-55°) tropospheric NO2

January July

Comparison of NASA "Standard Product" (SP; blue, red) and DOMINO v2 (green) NO2, based on same initial NO2 slant columns: differences lie in the separation of stratosphere and troposphere; SP2 is new w.r.t. SP1 and matches DOMINO better [Bucsela et al., 2013]

U = Usa E = EuropaA = East Asia

U E A

NO2 columns – NASA and KNMI

Page 9: OMI validation status

Three (sub)urban MAX-DOAS instruments against daily OMI tropospheric NO2 (x 1e15 molec/cm2) on some days in 2006-2011: magnitudes match, correlation not strong [Lin et al., 2013]

Monthly average tropospheric NO2 over Beijing from OMI and MAX-DOAS (2008-2011): OMI underestimates NO2 [Ma et al., 2013]

NO2 columns – KNMI

Page 10: OMI validation status

NO2 - NASA and KNMISO2 - NASA

NO2

SO2source: McLinden et al, 2013, ACPD

NO2-KNMI

NO2-NASA

SO2-NASA

improved AMF

original AMF

map column to in-situ via AQ model

Page 11: OMI validation status

Example for BeijingMa et al., ACP 2013

SCIA, 2008-2010 OMI, 2008-2009

MAX-DOAS: hourly averaged tropospheric NO2 VCDs during satellite overpass

red points: CTH < 1km; blue points: CTH > 1km

Comparison of OMI tropospheric NO2 VCD with (car) MAX-DOAS for different locations:

Beijing Wuxi New Delhi Paris Islamabad

OMI is always < MAX-DOAS

Page 12: OMI validation status

Validation of OMI satellite observations

Delhi-Agra, 16.01.2011Paris

R. Shaiganfar, MPIC Mainz

Shaiganfar et al., ACP 2011

Page 13: OMI validation status

Comparison of PDFs of effective cloud fraction

PDFs are practically the same. Not anticipated because of different sizes of the OMI and OMPS footprint.

13

eff. cloud fraction – NASAAlexander Vasilkov, SSAI

Page 14: OMI validation status

14

Southern mid-latitudes Tropics Northern mid-latitudes

cloud pressure – NASA and KNMI

Comparison of PDFs of cloud (optical centroid) pressure

- In general, OMI-NASA retrieves somewhat lower cloud OCPs than OMPS does. Differences are most pronounced in the tropics. - Differences between NASA and OMPS cloud pressures appear to be similar to differences between NASA and KNMI except for the differences in the tropics.

NASA KNMI OMPS

Alexander Vasilkov, SSAI

Page 15: OMI validation status

OMI BrO Measurements: Operational Data Analysis Algorithm & Initial Validation

OMI BrO – Data Consistency

Trend is varying from -0.7% per year at southern mid latitude to -1.7% per year at northern high latitude

Real or instrument degradation?

Raid Suleiman, CFA

Page 16: OMI validation status

OMI BrO Measurements: Operational Data Analysis Algorithm & Initial Validation

OMI BrO: Comparison with Ground-Based Observations

Prior to Row Anomaly OMI and Ground-Based Zenith-Sky total column BrO at Harestua, Norway.

Harestua columns by F. Hendrick (BIRA)

Comparison between OMI and Harestua (monthly mean) for the period of February 1, 2005 to August 8, 2012.

The varying trend here is very small

Raid Suleiman, CFA

Page 17: OMI validation status

OMI AOD (NASA) vs AERONET 44 sites, 4 years of data

Linear correlation coefficient: 0.81

linear regression (solid line): slope 0.79intercept: 0.10

[AP: agreement may be better than the authors suggest …..]

source: Ahn et al, 2014, JGR

Aerosol - NASA

Page 18: OMI validation status

ConclusionsSeveral OMI products have had some recent validation.

Uncertainties have been quantified and systematic effects characterized

since 2011: 18 new publications with validation results: 4 in 2012 (NO2, clouds, BrO, SO2), 9 in 2013 (aerosol, NO2, SO2), 2 in 2014 (aerosol, strat/total NO2), 1 submitted (aerosol), 2 in preparation (NO2, O3)

Gaps in the validation have been identified for almost all products. Main issues:

latest versions of products need validation: cloud(KNMI), HCHO, BrO, OClO

many products need characterisation over snow/ice or partially cloudy conditions

Most README files have been updated in the last 3 years; README files for HCHO and SO2 are more than 5 years old