Atmospheric Spectroscopy Lab UMBC

Anomaly Retrieval Notes + Other Tile Algorithms

Thu, Oct 27, 2022 L. Larrabee Strow

1. Telecon with Ryan: Oct 27, 2022

1.1. Overview of Compute Status

  • No data
  • Waiting: 9 months now for NASA to provide Cloud access for formatting CHIRP into tiles
  • Hopefully will start in ~1 month
  • For now concentrate on AIRS L1c tiles produced and stored at UMBC

1.2. AR Algorithm Status

  • 16-day anomaly retrieval development on hold (AIRS and CHIRP)
  • Instead concentrating on 20-year trends
    • For publication of various results
    • And to serve as validation for 16-day anomaly retrievals

1.2.1. Trend Liens

  • Using hottest quantiles as a proxy for clear
  • How much cloudiness do we introduce and are our Jacobians good enough?
  • Could our sampling be tested with other data?
  • Long term need MODIS cloud statistics info co-located with AIRS footprint (this is at least a year or more off)
  • Soon to be fixed: hot quantiles corrected for:
    • water vapor continuum effects (code done and run once)
    • use higher resolution emissivity database (I interpolated CAMEL to 3x5, should leave in 0.05 x 0.05 before 16-day 3x5 averaging of data)
  • I think we should try to do "all-sky" retrievals as well

1.2.2. 16-day AR Liens

  • Same sampling issues as above
  • Can we "quit" with say, 50% clouds?
  • Need to configure how to get 16-day Jacobians
    • From Merra2
    • Note: these Jacobians will have seasonal removed before use
    • Do each grid point for 16 days? 64 lats * 72 lons * 500+ 16 day time steps. I estimate this is ~1TB of data.
    • Or, come up with some kind of look-up table, machine-learning approach. Probably a lot of work…
    • Past experience indicates problems come up during ENSO events when averaged cloud anomalies just change too much

1.3. Targeted Papers

  • Trend work Sergio has been doing
    • Include Tsurf, T(z), H2O(z), etc?
    • Intercompare to ERA5, MERRA2, AIRS V7, CLIMCAPS, AMIPS?
  • Use AR retrievals to create time series of OLR by atmospheric variable
  • Quick paper concentrating soley on radiance trends??

1.4. Other Algorithms

  • Cloud forcing
  • Cloud fraction
  • Extreme events
  • Fires
  • Atmospheric variability on time

1.5. NASA HQ and Sounding

I real problem. What future do we want, etc.