Modelled and measured ammonia concentrations over The Netherlands by csgirla

VIEWS: 10 PAGES: 19

More Info
									National Institute
for Public Health
and the Environment
                      Modelled and measured ammonia
                      concentrations over The Netherlands - The
                      effect of dry deposition parameterization
                      Miranda van den Broek, Addo van Pul, Hans van
                      Jaarsveld (MNP), Roy Wichink Kruit
 The problem – The Dutch “ammonia gap”:
 Model concentrations underestimate measurements
 by ~30%:
                                                                  NH3                                    Observations:
                                                                                                         AMOR denuder, LML
                                         12.00                                                           network (8 stations
                                                                                                         throughout NL)
                      conc. ( ug m-3 )




                                          8.00
                                                                                                         Model:
                                          4.00                                                           - OPS source-receptor
                                                          measured                                       model [van Jaarsveld,
                                                          modelled                                       2004]
                                          0.00
                                                                                                         - Emissions from LEI
                                                 1994

                                                        1996

                                                               1998

                                                                      2000

                                                                             2002

                                                                                    2004

                                                                                           2006

                                                                                                  2008



National Institute
for Public Health
and the Environment
                        Overview
                 • The ‘ammonia gap’: problem description and possible
                   causes that were determined in earlier research (-> one of
                      them is modeled dry deposition)

                 • Description of the OPS model and its treatment of dry
                   deposition

                 • Simple approach: How much dry deposition (or: which Rc or
                   canopy resistance) would ‘close the gap’

                 • Does this Rc coincide with known uncertainties in the model
                   and with Rc derived from measurements of dry deposition?




National Institute
for Public Health
and the Environment
                      The possible causes:
                      • Observations – The observations in the Netherlands
                        LML (air quality measurements network) are carried out
                        with AMOR denuder systems. Comparisons with other
                        measurement technique shows no systematic difference
                        on longer (> day) time scales.

                      • Emissions – The underestimations are found especially
                        during the season of manure spreading (Feb – Apr, and
                        late summer) => VELD-study (also: see poster 1.1 by
                        Berkhout et al.)

                      • Dry deposition – Model parameters used for the
                        calculation of dry deposition may contain both
                        uncertainties and erroneous assumptions


National Institute
for Public Health
and the Environment
 The VELD-study (Achterhoek, NL), 2002-2003

                                                                                   (a)
                      40

                                                    Modeled
                                                   berekend    with OPS-KT

                      30
                                                    Measured (pass.samplers)
                                                   gem eten (gebiedsgem .)




                      20




                      10




                       0
                       13/A ug   10/Sep   8/Oct   5/No v   3/Dec   31/Dec 28/Jan 25/Feb 25/M a 22/A pr   20/M a   17/Jun   15/Jul   12/A ug




National Institute
for Public Health
and the Environment
                                      Components related to NH3

                                                    NH4 aerosol                                                                     NHx w et deposition


                               3.00                                                                                     800.00




                                                                                          deposition ( mol ha-1 a-1 )
                                                                                                                        600.00
              conc. (ug m-3)




                               2.00
                                                                                                                        400.00
                               1.00
                                              measured                                                                  200.00          measured
                                              modelled                                                                                  modelled
                               0.00                                                                                       0.00
                                      1994

                                             1996

                                                       1998

                                                              2000

                                                                     2002

                                                                            2004

                                                                                   2006




                                                                                                                                 1994

                                                                                                                                        1996

                                                                                                                                               1998

                                                                                                                                                      2000

                                                                                                                                                             2002

                                                                                                                                                                    2004

                                                                                                                                                                           2006
                                             = Problem is related to mass of ammonia, not to dispersion
                                             = comparison of other low altitude sources, e.g. NOx, agree well
                                             with measurements (~10% overestimate)
National Institute
for Public Health
and the Environment
                       NH3 in the OPS model



                                    Transport/            NH3              Chemical       NHx
                                                                          conversion
                                     dispersion       Concentration                       Conc. in air
                                                         in air
                      evaporation




                                                                emission
                                                              (not in OPS)


                                                           Dry                NHx dry         NHx wet
      NH3 emission                                      deposition           deposition      deposition


                                        - Plume model combined with trajectory model
                                        - All source-receptor relations are treated individually
                                            => non-linear processes such as re-emission cannot be included directly
                                        - Meteorological input: either hourly (OPS-KT) or through meteo-
National Institute
for Public Health
and the Environment                     classes (stability, wind direction, source-receptor distance)
                                        (standard OPS-LT)
                      DEPAC dry deposition module in OPS
                                            Fdep = Vd * ([NH3]air –[NH3]surf)

                                            Vd = 1/ (Ra + Rb +Rc)

                                            ----------------------------------------

                                            Ra Aerodynamic resistance


                                            Rb Quasi-laminar or boundary
                                            layer resistance


                                            Rc Canopy or surface
                                            resistance, with 3 deposition
                                            pathways:
                                                     - stomata (Rstom)

National Institute
                                                     - cuticulae (Rext)
for Public Health
and the Environment



                                                     - soil/ in canopy (Rsoil)
                      Errors/ uncertainties in DEPAC
                 • The modeled Rc ‘contains’ 3 deposition pathways, but it is derived from
                   measurements that may actually represent only one (total) or two (stomatal
                   vs cuticular/soil) pathways

                 • The land use class of the deposition measurements (wood, heathland) that
                   are the basis of DEPAC does not agree with the main land use class (=
                   fertilized grassland) in the model

                 • Strongly restricted deposition or even re-emission, known to occur
                   especially under warm/dry circumstances, is not included in the model

                 • Other parameterizations exist, based on other measurements of deposition
                   velocity/ resistance (but not necessarily better or worse)


                         All are leading to an underestimation of Rc


National Institute
for Public Health
and the Environment
      A simple exercise: Rc that “closes the gap”?
                                     NH3          NH4+          Wet dep.
                                     R2     regr. R2       Regr R2  Regr
                                                           .
                      Rc from        0.95   0.75   -0.32   0.72   0.70   0.67
                       DEPAC


                      Rc=RcDEPAC*2   0.93   0.86   -0.51   0.85   0.70   0.84

                      Rc=50          0.94   0.82   -0.56   0.80   0.70   0.76

                      Rc=100         0.93   0.96   -0.72   0.95   0.69   0.94

                      Rc=175         0.92   1.12   -0.79   1.11   0.69   1.06

                           Next: How do these values agree with
National Institute
for Public Health
and the Environment
                            known uncertainties and deposition
                                     measurements?
                      Note: only 2004 was examined in this test
                                               Ammonia-’gap’ for the past 15 years


                                                                      NH3


                                             12.00
                          conc. ( ug m-3 )




                                              8.00


                                              4.00
                                                              measured
                                                              modelled
                                              0.00
                                                     1994

                                                            1996

                                                                   1998

                                                                          2000

                                                                                 2002

                                                                                        2004

                                                                                               2006

                                                                                                      2008
National Institute
for Public Health
and the Environment
                                                 Effect of uncertainties in Rc

                                                          [NH3] 2004
                                                                                 x=y
                                            20
                                                                                 LML vs OPS with DEPAC
                      [NH3] (ug/m3) model




                                            15
                                                                                 LML vs "Rc=Rext"
                                            10
                                                                                 LML vs "Rc oneind.bij T>23"

                                            5                                    LML vs "Rc uit EMEP"

                                            0
                                                 0    5       10      15    20

                                                     [NH3] (ug/m3) meting


National Institute
for Public Health
and the Environment
      Effect of parameterization over multiple years



                                           2002         2003          2004          2005
                                          R2   regr   R2     regr   R2     regr   R2       regr
                      Standard        0.88 0.71       0.93   0.61 0.99     0.72   0.98     0.66

                      Rc = Rext       -        -      0.93   0.70 0.98     0.78        -      -

                      Rc:∞ at T>23º   -        -      0.93   0.73 0.98     0.83        -      -

                      Rc:∞at T>20º    -        -      0.93   0.77 0.97     0.86        -      -

                      Rc from EMEP 0.92        0.94   0.93   0.78   0.97   0.94   0.98     0.87
                             With EMEP module: “gap” averaged over these 4 years is
National Institute
                                           12% (instead of 32%)
for Public Health
and the Environment
                                    Difference between EMEP/DEPAC

                                                         Rext NH3 above grassland

                                    250


                                    200


                                    150
                      Rext (s/m )




                                                                                             DEPAC winter night/summer day
                                                                                             Rext from EMEP
                                    100                                                      DEPAC summer night/winter day



                                    50


                                     0
                                          20   30   40   50     60      70   80   90   100
                                                         relative humidity


National Institute
for Public Health
and the Environment
                                   Difference between EMEP/DEPAC
                                                        Rext NH3 above grassland

                                   250


                                   200

                                                                                            DEPAC winter night/summer day
                      Rext (s/m)




                                   150                                                      Rext from EMEP
                                                                                            DEPAC summer night/winter day
                                   100                                                      Rext meas. unfertilized grassl.
                                                                                            Rext meas.fit

                                   50


                                    0
                                         20   30   40   50     60      70   80   90   100
                                                        relative humidity




National Institute
for Public Health
and the Environment
                                         Rext measurements (2004-2006) at Haarweg station,
                                         Wageningen, the Netherlands (non-fertilized grassland)
                        Total Rc [s/m] compared with measurements
                                          nov.-jan.            feb-apr.             may-oct.
                      Rc at night         35                   100                  80
                      Rc at daytime       65                   105                  95
                      *Note:   only deposition cases are taken into account for these average Rc’s
 From Plantaz, PhD, measurements fertilized grassland (no recent manure spreading), Zegveld, ’92-’94

                                          nov.-jan.            feb-apr.             may-oct.

                      Rc at night         41                   64                   90

                      DEPAC               7                    7                    7

                      EMEP                36                   39                   50

                      Rc at daytime       71                   45                   78

                      DEPAC               7                    35                   34

                      EMEP                55                   112                  109

National Institute
for Public Health
and the Environment
                         From Wichink Kruit, measurements non-fertilized grassland, 2004-2006
                      Conclusions
 • A constant Rc of ~100-150 s/m would be able to “explain” the ammonia gap
 • This value is too high in the light of known uncertainties in DEPAC:
              - Treating the cuticular or external pathway as the total Rc, (no soil or stomatal
                path present) reduces the gap by ~8%pts.
              - Extra inhibited deposition at T> 23º or T>20º C reduces the gap by another 4%
                and 7% respectively (Note: still no emission was taken into account, so there
                may a possible ‘left-over’ uncertainty here!)
              - Using the EMEP routine instead of DEPAC reduces the gap by ~20%pts. That
                effect is comparable to the effect of a constant Rc of ~100 s/m
 • In measurements also Rc is somewhat lower than 100-150 s/m. On fertilized
   grassland, an average Rc was found that is comparable to the EMEP module.
   Again: emission data were left out
 • Nighttime Rc from EMEP (Rext) is higher than measured Rext on non-fertilized
   grassland


National Institute
for Public Health
and the Environment
                      Still work to be done...
         • Comparison of model results with measurements on a higher
           time resolution (e.g. diurnal cycle / seasonal values), for a
           better understanding of the variations in dry deposition.
           Detailed emission data are needed for this!

         • Get more detail in the treatment of compensation point and
           possible re-emission in the model.

         • From the Haarweg measurements: Rext is also related to
           [NH3]air concentration (see also Jones et al. [2001], Fowler et
           al., [1998], presentation by E. Nemitz here) -> This should also
           be included in the model.

         • Deposition measurements above arable land (the other
           landuse class that determines the NL NH3 mass balance…)


National Institute
for Public Health
and the Environment
                      The message to policy:

                      Model overestimation
                        of dry deposition
                       “closes the gap” at
                          least halfway

National Institute
for Public Health
and the Environment

								
To top