Evaluating radon-derived mixing depth as a potential length scale for nocturnal mixing processes over land S. Chambers, A.G. Williams, W. Zahorowski and A.D. Griffiths Australian Nuclear Science and Technology Organisation, Locked Bag 2001, Kirrawee DC, NSW 2232 Introduction To evaluate, and ultimately improve, numerical schemes for vertical mixing and exchange within the daytime and nocturnal boundary layers it is necessary to quantify mixing processes within the lower atmosphere at a temporal resolution sufficient to resolve the diurnal cycle. One way to quantitatively characterize near-surface mixing on diurnal time scales is to make continuous, high temporal resolution vertical gradient measurements of a suitable atmospheric tracer. Radon-222: a Passive Tracer Radon is a natural, radioactive, noble gas that is poorly soluble. Its terrestrial flux is relatively uniform and its only sink is radioactive decay. Radon’s 3.8-day half-life is ideal for atmospheric mixing studies, being larger than turbulent timescales (<1 hour) but short enough to ensure that concentrations in the free troposphere are typically much less than near the surface. Under strongly stable conditions, when the nocturnal mixing depth can become too shallow to be resolved by SODAR or LIDAR, near-surface radon concentrations remain intimately linked to local mixing. Measurement Site Hourly radon measurements have been made at 4 sites in the Sydney Basin since late 2006. At the Lucas Heights site a pair of radon detectors is used to measure the radon gradient between 2 and 50m. Here we summarise the first 2.5 years of Lucas Heights gradient observations. Network of four radon monitoring sites across the Sydney Basin. The Lucas Heights and Richmond sites lie within the Sydney airshed. 18km Lucas Heights 50m tower 1 km Lucas Heights is a topographically complex site, with changes in elevation of 150m within 1km of the tower. Location of site with respect to NSW coastline. Being <20km from the coast, this site is frequently influenced by maritime air masses. Lucas Heights Scale Hunter airshed Sydney airshed Illawarra airshed ANSTO Radon Detectors ANSTO has designed and built a suite of dual flow loop, two filter radon detectors which operate as follows: Sample air is passed through a delay volume to remove thoron ( 220 Rn), then through the 1 st filter, which removes ambient radon and thoron progeny. The filtered air is circulated within the delay volume until a portion of the sampled radon decays. The newly formed radon progeny are caught on the 2 nd filter. Their subsequent alpha decays are counted with a Zinc- sulphide scintillator and photomultiplier tube assembly. Data are stored on a Campbell Scientific data logger and transferred half hourly to a nearby computer. Schematic showing the principle of operation of the dual flow loop, two filter radon detector. A detector’s sensitivity is closely related to the volume of its delay chamber. Detectors range in size from 100 L to 5000L. The 5000 L detector can sense as few as 5 atoms of radon per litre of air. Measurements for this study were made using 1500 L radon detectors, with a flow rate of 100 L min -1 , response time of 45 minutes, sensitivity of ~0.3 counts per second per Bq m -3 of radon, and a lower limit of detection of ~40 mBq m -3 . Lucas Heights meteorological tower and radon detectors th e fi exhaust internal blower mixing volume radon delay volume ZnS scintillator signal to data logger second filter (wire screen) PMT and HV discriminator High flow internal flow loop Low flow external flow loop Sample In 220Rn Delay External Blower 1st Filter -50 -40 -30 -20 -10 100 110 120 130 140 150 160 170 Overland fetch -50 -40 -30 -20 -10 100 110 120 130 140 150 160 170 Marine fetch Summer fetch example. 3-hourly trajectories from 3 days of onshore flow in December 2007. Autumn / winter fetch example. 3- hourly trajectories from a persistent 3- day continental fetch event in May 2007. Seasonal Cycle of Radon Monthly radon distributions (10 th /50 th /90 th percentiles) at Lucas Heights show a pronounced seasonal cycle at 2m (grey) and 50m (green): characterised by a summer minimum and autumn / winter maximum. Seasonal change in air mass fetch, not the strength of the terrestrial radon flux, is the main driver of the observed seasonal cycle. NOAA HYSPLIT back trajectories show that the site fetch is primarily oceanic in summer and terrestrial in winter. Latitude Diurnal Cycle of Radon The composite diurnal cycle of radon at Lucas Heights is characterised by a maximum at sunrise and a late afternoon minimum. The amplitude of the diurnal radon signal at Lucas Heights was typically 8 times smaller than for Muswellbrook, a flatter site, 130km inland. The low amplitude diurnal signal at Lucas Heights is a combined result of the sites’ proximity to the coast and the locally complex topography. At night, under stable conditions, much of the locally emitted radon drains into the nearby valleys. 284 286 288 290 292 294 296 0 2 4 6 8 gradient Radon-222 (Bq m -3 ) Day of 2009 0 4 8 12 Wind speed (ms -1 ) 7 14 21 28 35 2m 50m Strong Rn gradients Weak Rn gradients Low Wind Speeds High Wind Speeds Strongly Stable Temperature ( o C) Near Neutral Hourly Radon Gradient Observations On hourly timescales the Lucas Heights radon gradient was highly variable, depending on the prevailing meteorology. Daytime gradients were small (typically <200 mBq m -3 ) due to strong convective and mechanical mixing. Nocturnal radon gradients were largest on relatively calm nights (little mixing), when near-surface temperature gradients were also large and positive. On the most strongly stable evenings, the 50m radon time series indicates that the depth of the stable nocturnal boundary layer (SNBL) was less than the height of the tower. Characteristics of the Nocturnal Radon Gradient The 942 days of hourly observations were binned according to the strength of the maximum nocturnal radon gradient: Strongly stable: ΔRn ≥ 1000 mBq m -3 Weakly Stable: 300 ≤ΔRn < 1000 mBq m -3 Near Neutral: ΔRn < 300 mBq m -3 On strongly stable evenings the 50m radon data stays close to the previous afternoon’s minimum, indicating that the SNBL depth is <50m. At sunrise, the near-surface radon is mixed upwards as the nocturnal inversion erodes. On weakly stable evenings the SNBL is deeper than 50m, and the radon concentrations increase with time at both heights. However, the stable stratification inhibits mixing, so the 50m radon signal is both damped and shifted compared to the 2m data . On strongly stable nights (large radon gradients), the corresponding 10m wind speed was <1 ms -1 and temperature gradients were in excess of 0.6 o C. On all of these occasions the 50m radon signal indicates that the height of the SNBL is <50m. Only a small increase in the 10m wind speed was required to lift the height of the SNBL above 50m (weakly stable cases). However, wind speeds in excess of 2.5 ms -1 were required to mix radon well in the lowest 50m (near neutral cases). The turbulent kinetic energy (TKE) and Bulk Richardson Number are alternative means by which to judge the mechanical and thermodynamic stability of the three radon- defined stability classes. Composite nocturnal radon gradients for the above three stability classifications. Composite hourly radon concentrations at 2m (filled symbols) and 50m (open symbols) for the weakly and strongly stable evenings. Composite TKE (from sonic anemometer at 10m) for strongly stable, weakly stable and near neutral conditions. TKE drops to ~0.1 m 2 /s 2 on strongly stable nights. Composite Bulk Ri for strongly stable, weakly stable and near neutral conditions. On stable nights when the wind direction is from the SE (oceanic fetch; local radon sources only), the build-up of radon at 2m between sunset and sunrise is steady and almost linear. Radon Derived SNBL Length Scale Under stable conditions, assuming a uniform surface radon flux, near-surface radon concentrations are closely linked to an effective mixing depth (h rad ) for the SNBL. Consider: m rad rad s m rad h h dC h dh F Ch C F dt dt λ + ⎛ ⎞ = − + − ⎜ ⎟ ⎝ ⎠ where C m is the mean SNBL Rn concentration, C h+ is the Rn concentration above the nocturnal inversion and λ is the Rn decay constant. Column- integrated radon budget Time change in radon Effective surface flux Radon decay Encroachment Entrainment ( ) ( ) ( ) 0 1 t t s rad m h F e h C C λ λ − − + − = − At night, knowing F s and assuming negligible entrainment, the Rn budget can be solved for h rad : Schematic of nocturnal radon profile (t0 = sundown; Fs=8.2 mBq/m 2 /s) Latitude SNBL depths predicted by the Australian Bureau of Meteorology’s Regional LAPS model. The large difference between h rad (~250m) and h LAPS (~50m) values on strongly stable nights is likely to be attributable to a loss of locally-emitted radon to valley drainage flows ie. F s (effective) << F s (measured). LAPS does not resolve the local topography 2 and 50m composite radon concentrations on strongly stable nights under oceanic, mixed and terrestrial fetch conditions. Diurnal composite radon-derived mixing depths on strongly stable nights under oceanic, mixed and terrestrial fetch conditions. Reconstructed normalised nocturnal radon profiles for stable and near neutral conditions. Measurements have been averaged into altitude bins Δ(z/hrad) = 0.05. 0 1000 2000 3000 4000 5000 6000 Jan Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec Month of composite year Radon (mBq m -3 ) 0 3000 6000 9000 12000 15000 0 2 4 6 8 10 12 14 16 18 20 22 Hour of composite day Muswellbrook Radon (mBq m -3 ) 0 1000 2000 3000 4000 5000 Lucas Heights Radon (mBq m -3 ) 0 1000 2000 3000 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 Hour of composite day Radon gradient (mBq m -3 ) Sunrise 0 1000 2000 3000 4000 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 Hour of composite day 2 and 50 m Radon (mBq m -3 ) Sunrise 0.0 0.3 0.6 0.9 1.2 1.5 1.8 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 Hour of composite day TKE (m 2 /s 2 ) Near neutral -1.0 -0.5 0.0 0.5 1.0 1.5 2.0 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 Hour of composite day Bulk Richardson Number Strongly stable 0 1000 2000 3000 4000 5000 18 20 22 0 2 4 6 8 18 20 22 0 2 4 6 8 18 20 22 0 2 4 6 8 Hour of composite day Radon (mBq m -3 ) SE Fetch, Oceanic S Fetch, Mixed SW Fetch, Terrestrial 0 300 600 900 1200 18 19 20 21 22 23 0 1 2 3 4 5 6 7 8 9 Hour of composite day h LAPS (m agl) Strongly Stable Near Neutral Monthly radon distributions Composite diurnal cycle ( ) 2 i v g R U θ θ Δ = Δ 0 1 h m C Cdz h = ∫ 0.00 0.10 0.20 0.30 0.40 0.50 0.00 0.10 0.20 0.30 0.40 0.50 0.60 0.70 Rn(50) / Rn(2) z(50) / h rad h/L>10 (Stable) h/L<2 (near neutral) 0 200 400 600 20 21 22 23 0 1 2 3 4 5 6 7 Hour of composite day h rad (m agl) SE fetch S fetch SW fetch