ABL structures and processes at Sodankyl Carl Fortelius

ABL structures and processes at Sodankylä Carl Fortelius HIRLAM All Staff Meeting, Cracow, 13 -16 April 2010 27/11/2020

Contents. . . • Intercomparison of measurements and two weather forecasting models, HIRLAM 7. 2 and 7. 3 beta 2 • Descriptive approach: looking for differences and similarities 27. 11. 2020 2

Contents. . . • ABL structures • Surface layer temperature gradient • winter-time cold spells, summer-time diurnal cycle • Surface layer wind shear • too few cases of low wind speed at 10 m (0 -2 m/s, say), e. g winter and night time summer • and processes • Radiative and turbulent heat fluxes • At sodankylä • Sodankylä (67 N, 26 E) • Sparse Scots pine forest, 10 -15 m tall • July 2009, Dec 2009 -Jan 2010 27. 11. 2020 3

Contents. . . • Models • HIRLAM operational reference runs (RCR) • HIRLAM 7. 2 • HIRLAM 7. 3 beta 2 • Dec 2009 -Jan 2010: RCR-parallel run at FMI • July 2009: as RCR, but with 3 D-VAR and without LSMIX • state at h+6, fluxes during hours 0. . . 6 27. 11. 2020 4

Contents. . . • Data • The Arctic Research Centre of the Finnish Meteorological Institute (FMI-ARC) • T: 3 m, 30 m • V: 18 m, 38 m • Rad: 18 m • SHF, LHF: 22 m • Soil heat flux: -7 cm 27. 11. 2020 5

Temperature more cold cases in new model Sodankylä Dec 2009 -Jan 2010 Yellow: Obs Red: RCR Blue: 7. 3 beta 2 more cold bias in new model 27. 11. 2020 6

Surface layer temperature gradient WINTER Sodankylä Dec 2009 -Jan 2010 Black: Obs Red: RCR Blue: 7. 3 beta 2 27. 11. 2020 7

Surface layer temperature gradient Sodankylä Dec 2009 -Jan 2010 Black: Obs Red: RCR Blue: 7. 3 beta 2 Inversions 27. 11. 2020 8

Inversions and surface fluxes Net radiatin vs. stability Turbulent heat flux vs stability 27. 11. 2020 9

Inversions and surface fluxes few good inversion cases covered by the observations 27. 11. 2020 10

Inversions and surface fluxes both models produce negative net radiation but only 7. 3 generates inversions 27. 11. 2020 11

Inversions and surface fluxes too much downward turbulent heat fluxes during inversions 27. 11. 2020 12

Ensemble mean surface heat fluxes SHF Soil LHF Rad 27. 11. 2020 13

Ensemble mean surface heat fluxes SHF Soil LHF Rad Too much radiative cooling 27. 11. 2020 14

Ensemble mean surface heat fluxes SHF Too much downward sensible heat flux in RCR, better in 7. 3 Soil LHF Rad Too much radiative cooling 27. 11. 2020 15

Ensemble mean surface heat fluxes SHF Soil LHF Rad Too much downward sensible heat flux in RCR, better in 7. 3 Too much heat release from the soil (residual term) Too much radiative cooling 27. 11. 2020 16

Average temperature Dec 2009 -Jan 2010 (diurnal cycle) screen 30 m 27. 11. 2020 17

Average temperature Dec 2009 -Jan 2010 cold bias, screen 30 m 27. 11. 2020 18

But, for whole Finland, 7. 2 is cold at screen level, 7. 3 is ok 27. 11. 2020 19

Surface layer temperature gradient SUMMER Sodankylä July 2009 Black: Obs Red: RCR Blue: 7. 3 beta 2 27. 11. 2020 20

Surface layer temperature gradient SUMMER Sodankylä July 2009 Black: Obs Red: RCR Blue: 7. 3 beta 2 shape better in 7. 2 27. 11. 2020 21

Surface layer temperature gradient Sodankylä July 2009 Black: Obs Red: RCR Blue: 7. 3 beta 2 very unstable 27. 11. 2020 22

Temperature gradient and heat fluxes Net radiation vs stability Turbulent fluxes vs, stability 27. 11. 2020 23

Temperature gradient and heat fluxes similar fluxes, stronger gradient Net radiation vs stability Turbulent fluxes vs, stability 27. 11. 2020 24

Ensemble mean surface heat fluxes Rad Soil LHF SHF 27. 11. 2020 25

Ensemble mean surface heat fluxes Rad Less radiative heating Soil LHF SHF 27. 11. 2020 26

Ensemble mean surface heat fluxes Rad Less radiative heating Soil LHF Opposite relative partitioning into SHF and LHF SHF 27. 11. 2020 27

Ensemble mean surface heat fluxes Rad Observed budget not closed: Rad+SHF+LHF = 20 W/m 2 soil? Less radiative heating Soil LHF Opposite relative partitioning into SHF and LHF SHF 27. 11. 2020 28

Temperature diurnal cycle screen 30 m 27. 11. 2020 29

Temperature diurnal cycle cold bias screen 30 m 27. 11. 2020 30

Temperature diurnal cycle cold bias screen 30 m no night time bias 27. 11. 2020 31

But, for whole Finland, biases are small: 18 16 14 27. 11. 2020 32

Surface layer wind shear Yellow: obs Red: RCR Blue: 7. 3 beta 2 27. 11. 2020 33

Surface layer wind shear Yellow: obs Red: RCR Blue: 7. 3 beta 2 ”Data shift, models stay the same” 27. 11. 2020 34

Surface layer wind shear vs stability Black: obs Red: RCR Blue: 7. 3 beta 2 27. 11. 2020 35

Surface layer wind shear vs stability Black: obs Red: RCR Blue: 7. 3 beta 2 too little shear in RCR 27. 11. 2020 36

Surface layer wind shear vs stability Black: obs Red: RCR Blue: 7. 3 beta 2 too little shear in RCR too much shear in 7. 3 27. 11. 2020 37

U 10 m, diurnal cycle, July 2009, Finland 4 m/s OBS 7. 2 3 m/s 7. 3 beta 2 27. 11. 2020 38

Summary 1: temperature and energy cycle • Winter • New HIRLAM: more frequent and more intense surface incersions • New HIRLAM: Increased cold bias at the top of the surface layer • Both versions overestimate the intensity of the surface energy cycle (too much radiative loss, too much downward sensible heat flux, too much heat release from the soil). Some improvement in the new version. • Summer • Both versions underestimate the radiative heating • Observed SHF exceeds observed LHF in magnitude, the opposite is true for both models • Both models show a cold day-time bias throughot the sfc layer 27. 11. 2020 39

Summary 2: wind • The ratio V(10 m)/V(NLEV) • HIRLAM 7. 3 overestimates shear in unstable situations • HIRLAM 7. 2 underestimates shear in neutral to stable situations • The observed distribution shows stronger annual cycle than either of the model versions 27. 11. 2020 40

Distributions of wind speed 27. 11. 2020 41