NACP Interim Synthesis: Continental Fluxes from Forward and Inverse Models

Andy Jacobson, Mac Post, Debbie Huntzinger, Bob Cook, and lots of modelers

2

Atmospheric Inversions

ForwardModels

Flux tower sites

NEE

flux footprint

size

site

continent

3

North AmericaBoreal (NABR)

North America Temperate (NATM)

4

5

6

7

8

9

10

11

12

Inversions25th

percentileCentral

75th percentile

Uptake 0.7 1.0 1.4

IAV peak-peak 0.8 1.2 1.6

IAV (sd) 0.26 0.34 0.44

Forward models

25th percentile

Central75th

percentile

Uptake 0.1 0.3-0.5 0.7

IAV peak-peak 0 0.2 0.4

IAV (sd) 0.04 0.10 0.19

North America (boreal and temperate) summary stats for NEE

13

Atmospheric Inversions

ForwardModels

MODIS

Flux tower sites

NEE GPP

14

15

16

17

No apparent relationship between GPP and NEP(in the same year, anyway)

18

Conclusions

Inversions do manifest more seasonality, uptake, and IAV over N. America than forward models.

Inversions don't agree as well as one might hope.

MODIS GPP generally smaller than forward simulations.

MODIS, CASA models and DLEM see less GPP in 2002, but NEE signal is not apparent.

19

20

21