CLIMATOLOGY OF AIRSEA ENERGY EXCHANGE SW Sensible heat

CLIMATOLOGY OF AIR-SEA ENERGY EXCHANGE SW Sensible heat LW Latent heat

CLIMATOLOGY OF AIR-SEA ENERGY EXCHANGE Ocean surface heat balance: H = SW - LW - Qh - Qe 0 100 65 8 27 Net heat flux

CLIMATOLOGY OF AIR-SEA ENERGY EXCHANGE Ideally the globally integrated surface net flux should converge to zero. Uncertainties of the heat exchange through the ocean bottom and heat inflow/outflow with rivers and underground water are small. Major features of the net flux fields: Spatial patterns are more comparable with surface sensible and latent fluxes, which are not zonal in contrast to SW and LW radiation. How to produce the flux field? Ø Individual variable information Ø Correction of data, use of metadata Ø Selection of schemes Ø Computation of fluxes Ø Composition of the fields (averaging)

Basic surface meteorological parameters Sea surface temperature A general definition of sea surface temperature (SST) is that it is the water temperature at 1 meter below the sea surface. However, there a variety of techniques for measuring this parameter that can potentially yield different results because different things are actually being measured. The sea surface skin temperature (SSST) is the temperature that physically determines the surface heat fluxes. It may be measured radiometrically from ships and other in situ platforms, and by satellite-borne radiometers provided the atmospheric effects are properly corrected. The cooling due to sensible and latent heat fluxes and the longwave emission occurs at the skin, whereas the shortwave heating is distributed over a greater depth. Thus, most of the time, the SSST is colder than the water just beneath the skin, typically by a few tenths °C. This difference increases with increased surface cooling and decreases with increasing wind speed.

For traditional bulk formulae the transfer coefficients have been determined with respect to the bulk SST, so for application of these formulae, this would be a more appropriate temperature to use. For the bulk formulae which use the transfer coefficients derived from surface renewal theory the skin temperature is the appropriate value. Bucket SST measurements Infrared radiometer

Ship SST data are obtained mostly from Engine Room Intake (ERI) thermometers or (about 1/3 of the modern data) from SST buckets. A small but increasing number of ships use hull contact sensors which, if carefully calibrated, appear to give the most consistent SST data (Kent et al. , 1993 a; Emery et al. , 1997). ERI SST data are warmer under most conditions, on average by 0. 35 C although there is significant scatter about this typical value. Bucket measurements are found only to be biased compared to hull values during sunny daytime conditions when they gave on average SST values about 0. 3°C warmer. This is more likely due to the buckets heating on deck prior to use rather than to near surface ocean heating. VOS SST correction: All engine intake samples, when identified, should be reduced by 0. 35 C, otherwise - default reduction by 0. 2 C

Air temperature For atmospheric temperature and humidity, the most accurate is a psychrometer (wet and dry bulb thermometer) whose measurements are based on well-established thermodynamic theory. The most critical requirements to attain its potential accuracy are: Ø adequate ventilation of air past the sensing elements (3 -4 ms-1 flow rate) Ø to ensure the full wet bulb depression, and adequate shielding from solar radiation. This usually means a double shield with the space between also ventilated. Basic accuracy depends on the type of sensing element used; for the familiar sling and Assman psychrometers this is the precision of the particular mercury-in-glass thermometer, 0. 1°C at best.

The exposure of thermometer screens on the VOS varies from good (e. g. screens hung on stanchions on the outboard rails of either bridge wing) to very bad (e. g. "the screen is made of brown varnished wood and fitted to the side of the wheelhouse in the 'porch' of the bridge wing on the port side"). The poorly exposed sensors are about 0. 5°C warm. During the day all the sensors showed increasingly warm readings with increasing solar radiation. For the better exposed sensors this bias was up to 2°C; for the poorly exposed sensors the mean bias reached over 4°C. Air temperature correction: Ta = (2. 7 -0. 064 urel) SW / 1000

Humidity The wet and dry bulb psychrometer is the traditional meteorological instrument for the routine measurement of temperature and humidity. In general, however, psychrometers are not suitable for continuous routine measurement of atmospheric humidity at sea in stand-alone or automatic mode because of their need for attention (e. g. washing salt from the wick, replenishing the water reservoir). Dewpoint hygrometers are also based on sound thermodynamic theory, measuring the temperature at which a film of dew forms on a cooled mirror, but are generally too complex to serve as operational instruments. Their main use in air-sea studies is as a reference standard; accuracy of 0. 2°C in dewpoint is readily achievable (corresponding to 0. 2 g/kg-1 at about 22°C).

Humidity - more The VSOP-NA results showed that psychrometers produced lower (and therefore presumably more accurate) dew point readings compared to screens. Since the ship may often be a source of heat but is rarely a significant source of water vapour, shipboard humidity readings may be of better quality than the temperature data. Correction of the humidity (dew point temperature): For unaspirated screen measurements Tdew’ = 1. 029 Tdew - 1. 080, where prime denotes the corrected dew point temperature. Onethird of this correction is applied if no information is available about the method of humidity measurements

Wind speed 1. Anemometer winds In general are considered to be most accurate, if: Ø The anemometer is properly installed onboard the ship Ø True wind is properly computed from the relative wind (both requirements normally are not the case) 2. Beaufort estimates of wind Less accurate, but more homogeneous Flow distortion by the ship superstructure: RV Darvin Laboratory and numerical modeling can help identify biases in the measured wind speed. It is difficult to derive flow distortion effects for all ships. RV Knorr

Beaufort wind estimates Rear-Admiral, Sir Francis Beaufort, Knight Commander of the Bath, was born in Ireland in 1774. He entered the Royal Navy at the age of 13 and was a midshipman aboard the Aquilon. Beaufort is said to have had an illustrious career on the seas and by 1800 had risen to the rank of Commander. In the summer of 1805 Commander Beaufort was appointed to the command of the Woolwich, a 44 gun man-of-war. It was at this time that he devised his wind force scale. An early surviving form the scale is replicated below. By 1838 the Beaufort wind force scale was made mandatory for log entries in all ships of the Royal Navy. Beaufort last served as Hydrographer to the Admiralty. He died in 1857 two years after his retirement. In 1854 the English and French were entrenched in fighting at Sevastopool. The fleets carrying almost all their winter supplies was struck by an intense, early winter storm on the morning of November 14. In response to the losses and with the hope that there might be some way to forecast future storms, the British Admiralty and the French Marine jointly sponsored a weather network -- the ancestor of the World Meteorolgical Organization -- to provide storm warnings. And here then is when Sir Beaufort's scale begins its protean growth.

Figures to Denote the Force of the Wind 1 Light Air Or just sufficient to give steerage way. 2 Light Breeze 3 Gentle Breeze 4 Moderate Breeze Or that in which a man-of-war with all sail set, and clean full would go in smooth water from. 5 Fresh Breeze Royals, &c. Strong Breeze Single-reefed topsails and topgal. sail 6 7 Moderate Gale 8 Fresh Gale 9 Strong Gale Or that to which a well-conditioned man-of-war could just carry in chase, full and by. 1 to 2 knots 3 to 4 knots 5 to 6 knots Double reefed topsails, jib, &c. Treble-reefed topsails &c. Close-reefed topsails and courses. 10 Whole Gale Or that with which she could scarcely bear close-reefed main-topsail and reefed fore-sail. 11 Storm Or that which would reduce her to storm staysails. 12 Hurricane Or that which no canvas could withstand.

EQUIVALENT SCALES in m/s

Observational height corrections: Typical observational heights vary on different ships from the first meters to several tens of meters. On oil and drilling platforms observational heights can approach 6080 m. Corrections should be applied using the same bulk parameterizations which are expected to be used for the flux computations. Iteration scheme: Coefficients and flux – profile relationships for “uncorrected” heights Correction of winds, temperatures and humidity according to the derived profiles Re-computation of the coefficients and fluxes

Flux averaging and climatological fields: Averaging of the computed fluxes in the space-time coordinates may suffer from the number of observations (inadequate sampling). >500 <5

Nature of sampling bias in VOS fluxes: time dependent biases

Precipitation: The use of conventional rain-collecting instruments, designed for land use, results in uncertainties which are of the same order of magnitude as the mean precipitation estimates. q the effect of the flow around the ship’s overall structure which can lead to undercatch or overcatch depending on the location of gauge; q the effect of the flow in the close vicinity of the rain-gauge, which tends to carry the rainabove the orifice of the gauge and leads to a wind speed dependent undercatch. Parameterization of precipitation using the weather code information: Tucker (1961), Dorman and Bourke (1978): x=1. 85 mm, y=5. 66 mm, z=8. 13 mm (per 3 hours).

Surface fluxes from in-situ: VOS Woodruff et al. 2009 Reports which allow to compute the fluxes Gulev et al. 2007 1999: NOC 1. 1 (SOC) climatology: q 1979 -1993, monthly q variable corrections, WMO-47 q successive correction Global imbalance: 30 NOC-2. 0 mean net heat flux W/m 2 2009: NOC-2. 0 climatology: q q 1973 -2006, daily, updated variable corrections, WMO-47 fluxes from optimally interpolated variables uncertainty estimates for all fluxes Global imbalance: 24 W/m 2 Berry and Kent 2009 The problem of global closure in VOS still remains unresolved!

Satellite-based surface fluxes ISCCP Radiative fluxes (SW+LW): MODIS, TOVs, GOES global products, 2000+ global, 3 -hourly, 1983 -onwards ISCCP-FD SW Pinker et al. 2009 ISCCP-FD LW MPI-HOAPS GISS-GSSTE J-OFURO Turbulent fluxes from satellites: (Klepp et al. 2008) (Chou et al. 2004) (Kubota Low et al. 2007) Better accuracy SSM/I BEL (Bouras-Eymard-Liu 2002) IFREMER (Bentamy AMSU NSCAT ERS et al. 2008) Q-SCAT AVHRR SH ~ Ch (SST – Ta ) V LH ~ Ce (q 0 (SST) – qa) V Parameterizations (COARE, etc) HOAPS-3 latent heat, W/m 2

Satellite precipitation from SSM/I, AMSU, AQUA, TRMM missions GPCP products: 1°× 1°, daily, 1997+ 0. 25°× 0. 25°, 3 -hr, 2002+ GPCP daily precip, Jan 2008 CMORPH:

Old and modern era reanalyses

Surface fluxes in reanalyses SW radiation – bulk parameterizations or RTM LW radiation – RTM Qh, Qe, τ – bulk parameterizations as parts of boundary layer models Individual variables (Ta, ez, V) – prognostic, i. e. instantaneous at forecast time SLP, Precipitation - diagnostic – i. e. 6 -hourly averages prior the forecast time FLUXES – diagnostic – i. e. 6 -hourly averages prior the forecast time SST – diagnostic (boundary condition at forecast time) Ta, ez, V Qh, Qe, SW, LW, τ 6 hours

Surface fluxes from reanalyses: Short wave radiation: NCEP vs ISCCP-FD NCEP 1 ISCCP Trenberth et al. 2009, CWP NCEP 2 ISCCP W/m 2 ERA-Interim precipitation improvements Precipitation over tropical oceans in different mm/da reanalyses vs observations y et al. Improvement of fluxes in. Simmonds 2009 1989 1990 1991 1992 reanalyses and NWP is not a problem of better flux parameterizations per se: the whole system (physics, numerics, data assimilation) must be improved Operationalization and a new level of sustained funding

Surface fluxes from reanalyses: Trenberth 2011

Surface fluxes from reanalyses: Trenberth 2011

Is there a “fit-for-all” flux product? Regional balance s and MHT SH+LH, WHOI OA SH+LH, NCEP 2 Atlantic MHT Yu and Weller 2007 Variabilit y and trends 19482007 Gulev et al. 2007 correlati on Acceleration of the global water cycle: WHOI-OA evaporation vs GPCP precipitation