date: 2022-09-21
There are several instances across the atmospheric isotope measurements made by NEON where there appears to be a disconnect between the “known” isotope ratio (and in the case of CO2, the mixing ratio) of the reference materials used. NEONiso attempts to correct these to maximize data coverage, primarily through: a) analysis of time series plots of reference and measured values of validation gases, and/or b) consultation with NEON staff.
As of version 0.6.0, these corrections are only applied to carbon isotopes, as the water isotope calibration functions are still experimental and have known deficiencies.
The list of carbon isotope reference material corrections can be accessed using the ‘carb’ data object bundled with NEONiso:
#> # A tibble: 27 × 11
#> site refGas startDate co2_old co2_corr co2_repairedRaw d13C_old
#> <chr> <chr> <dttm> <dbl> <dbl> <lgl> <dbl>
#> 1 ABBY co2High 2019-10-03 00:00:00 509. NA FALSE -9.65
#> 2 CLBJ co2High 2020-01-09 00:00:00 508. NA FALSE -9.26
#> 3 GRSM co2High 2020-03-11 00:00:00 533. NA FALSE -13.0
#> 4 JERC co2High 2019-11-13 00:00:00 525. NA FALSE -10.4
#> 5 JERC co2High 2020-02-25 00:00:00 507. NA FALSE -9.53
#> 6 MOAB co2Low 2020-03-11 00:00:00 368. NA FALSE -12.4
#> 7 MOAB co2Med 2020-03-11 00:00:00 444. NA FALSE -9.06
#> 8 BARR co2High 2018-06-01 00:00:00 425. 555. FALSE -9.99
#> 9 BARR co2Low 2018-06-01 00:00:00 386. 359. FALSE -8.83
#> 10 BLAN co2Low 2019-04-01 00:00:00 361. 363. FALSE -8.54
#> 11 BLAN co2Med 2019-04-01 00:00:00 414. 436. FALSE -9.52
#> 12 BLAN co2High 2019-04-01 00:00:00 509. 521. FALSE -10.9
#> 13 ONAQ co2Low 2018-06-18 00:00:00 421. 367. FALSE -9.38
#> 14 ONAQ co2Med 2018-06-18 00:00:00 451. 412. FALSE -10.6
#> 15 ONAQ co2High 2018-06-18 00:00:00 472. 467. FALSE -13.0
#> 16 ORNL co2Low 2020-01-01 00:00:00 351 368. TRUE -8.6
#> 17 ORNL co2Med 2019-11-14 00:00:00 530. 443. FALSE -15.9
#> 18 ORNL co2High 2019-11-14 00:00:00 443. 530. FALSE -10.6
#> 19 SRER co2Low 2018-04-01 00:00:00 385 356. TRUE -9.15
#> 20 SRER co2Med 2018-04-01 00:00:00 490 429. TRUE -14.1
#> 21 STER co2Low 2019-01-16 00:00:00 351 360. FALSE -8.5
#> 22 TREE co2Low 2020-01-01 00:00:00 345 369. TRUE -8.9
#> 23 TREE co2Med 2020-01-01 00:00:00 395 438. TRUE -9.9
#> 24 TREE co2High 2020-01-01 00:00:00 525 511. TRUE -16.1
#> 25 WOOD co2Low 2018-06-18 00:00:00 363. 367. FALSE -8.83
#> 26 WOOD co2Med 2018-06-18 00:00:00 460. 413. FALSE -10.5
#> 27 WOOD co2High 2018-06-18 00:00:00 564. 455. FALSE -11.6
#> d13C_corr d13C_repairedRaw versionAdded
#> <dbl> <lgl> <dbl>
#> 1 NA FALSE 0.6
#> 2 NA FALSE 0.6
#> 3 NA FALSE 0.6
#> 4 NA FALSE 0.6
#> 5 NA FALSE 0.6
#> 6 NA FALSE 0.6
#> 7 NA FALSE 0.6
#> 8 -11.5 FALSE 0.4
#> 9 -8.61 FALSE 0.4
#> 10 -8.96 FALSE 0.4
#> 11 -10.1 FALSE 0.4
#> 12 -15.4 FALSE 0.4
#> 13 -9.05 FALSE 0.4
#> 14 -9.10 FALSE 0.4
#> 15 -10.5 FALSE 0.4
#> 16 -9.05 TRUE 0.4
#> 17 -10.6 FALSE 0.6
#> 18 -15.6 FALSE 0.6
#> 19 -8.70 TRUE 0.4
#> 20 -10.4 TRUE 0.4
#> 21 -8.76 TRUE 0.4
#> 22 -9.18 TRUE 0.4
#> 23 -10.3 TRUE 0.4
#> 24 -14.8 TRUE 0.4
#> 25 -9.27 FALSE 0.4
#> 26 -9.45 FALSE 0.4
#> 27 -10.4 FALSE 0.4
#> notes
#> <chr>
#> 1 <NA>
#> 2 <NA>
#> 3 <NA>
#> 4 <NA>
#> 5 <NA>
#> 6 <NA>
#> 7 <NA>
#> 8 <NA>
#> 9 <NA>
#> 10 <NA>
#> 11 <NA>
#> 12 <NA>
#> 13 <NA>
#> 14 <NA>
#> 15 <NA>
#> 16 <NA>
#> 17 issue first corrected in 0.4, updated for 0.6 - still looks suspect
#> 18 issue first corrected in 0.4, updated for 0.6
#> 19 <NA>
#> 20 <NA>
#> 21 <NA>
#> 22 <NA>
#> 23 <NA>
#> 24 <NA>
#> 25 <NA>
#> 26 <NA>
#> 27 <NA>This table lists, in order, a) the site affected; b) the reference gas affected; c) the start and end dates to which the correction is applied; d) the suspect CO2 reference mole fraction; e) the corrected CO2 mole fraction; f) logical of whether files recently pulled from the NEON data portal are expected to have the CO2 values already corrected; f) the suspect d13C value; g) the corrected d13C value; h) the version of NEONiso where that entry was added to the table; and i) any additional notes regarding the correction.
Note that because this table may change over time, it is possible that calibrations performed by different versions of the code may change over time (though, hopefully for the better.) Please post any suspected missing values to the maintainer as issues on the GitHub repo.