Roughness length and drag coefficient at two MONTBLEX-90 tower stations

Using MONTBLEX-90 mean velocity data, roughness lengths and drag coefficients are estimated at Jodhpur and Kharagpur. At Jodhpur, since the surface is not uniform the roughness length is estimated separately in three different subsectors within the range of prevailing wind directions and averages to 1.23 cm in the sector between 200° and 230° which is relatively flat with no obstacles on the ground. At Kharagpur, where the terrain is more nearly homogeneous, the average value (for all prevailing wind directions) is 1.94 cm. The drag coefficient C_D at Jodhpur shows variation both with the roughness subsector and with wind speed, the average over all directions increasing rapidly as themean wind speed Ū₁₀ at 10m height drops according to the power lawC _D = 0.05 Ū ₁₀ ^t-1.09 in trie range 0.5 < Ū₁₀ < 7 m s⁻¹. At Kharagpur, the drag coefficient is smaller than at Jodhpur by nearly 50% for the same range of wind speeds (> 3 ms⁻¹).     

An analysis of MONTBLEX data on heat and momentum flux at Jodhpur

Parameterization of sensible heat and momentum fluxes as inferred from an analysis of tower observations archived during MONTBLEX-90 at Jodhpur is proposed, both in terms of standard exchange coefficientsC _H andC _D respectively and also according to free convection scaling. Both coefficients increase rapidly at low winds (the latter more strongly) and with increasing instability. All the sensible heat flux data at Jodhpur (wind speed at 10 m Ū₁₀ < 8 ms⁻¹) also obey free convection scaling, with the flux proportional to the ‘4/3’ power of an appropriate temperature difference such as that between 1 and 30 m. Furthermore, for Ū₁₀ < 4 ms⁻¹ the momentum flux displays a linear dependence on wind speed.     

The value ofC efor the Arabian Sea during summer monsoon

We estimate, from the moisture budget the bulk aerodynamic coefficient for latent heat flux (C _e) during the monsoon season over the central Arabian Sea. The average value ofC _eunder active monsoon conditions was found to be 2.25 × 10⁻³ which is nearly 60% higher than those previously used.     

Benthic nutrient fluxes in the intertidal flat within the Changjiang （Yangtze River） Estuary

In an annual cycle from March 2005 to February 2006, benthic nutrient fluxes were measured monthly in the Dongtan intertidal flat within the Changjiang （Yangtze River） Estuary. Except for NH4^＋, there always showed high fluxes from overlying water into sediment for other four nutrients. Sediments in the high and middle marshes, covered with halophyte and consisting of macrofauna, demonstrated more capabilities of assimilating nutrients from overlying water than the low marsh. Sampling seasons and nutrient concentrations in the overlying water could both exert significant effects on these fluxes. Additionally, according to the model provided by previous study, denitrification rates, that utilizing NO3- transported from overlying water （Dw） in Dongtan sediments, were estimated to be from -16 to 193 μmol·h^-1·m^-2 with an average value of 63 μmol·h^-1·m^-2 （n=18）. These estimated values are still underestimates of the in-situ rates owing to the lack of consideration of DN, i.e., denitrification supported by the local NO3^- production via nitrification.     

A numerical simulation of residual circulation in Tampa Bay. Part I: Low-frequency temporal variations

The residual (time-average) salinity and circulation in a numerical ocean model of the Tampa Bay estuary are shown to experience significant temporal variation under realistic forcing conditions. A version of the Estuarine Coastal Ocean Model developed for Tampa Bay with 70 by 100 horizontal grid points and 11 sigma levels is examined for the years 2001–2003. Model output variables are averaged over the entire time of the simulation to generate long-term residual fields. The residual axial current is found to be dominated by the buoyancy-driven baroclinic circulation with an outflow (southwestward) at the surface and to the sides of the shipping channel, and an inflow (northeastward) usually occurring subsurface within or above the shipping channel. Averages over 30 d are used to examine variations in the residual fields. During the simulation the average surface salinity near the head of Tampa Bay varies with the freshwater inflow, from 12‰ to 33%. At the bay mouth salinity varies from 30%. to 36%.. A localized measure of the baroclinic circulation in the shipping channel indicates the residual circulation can vary strongly, attaining a magnitude triple the long-term mean value. The baroclinic circulation can be disrupted, going to near zero or even reversing, when the buoyancy-driven flow is weak and the surface winds are to the northeast. Three time periods, representing different environmental conditions, are chosen to examine these results in detail. A scaling argument indicates the relative strength of buoyancy versus wind as ΔρgH²(LC _Dω²)⁻¹, where δρ is head-to-mouth density difference across the bay,g is gravitational acceleration,H is depth,L is bay length,C _D is the surface wind drag coefficient, andw is wind speed. Tampa Bay is usually in the buoyancy dominated regime. The importance of winds in the weak-buoyancy case is demonstrated in an additional simulation without wind stress.     

Fluxes of heat and momentum over sea surface during the passage of a depression in the north Bay of Bengal

Time variation of surface fluxes of heat, moisture and momentum over a sea station (20°N 89°E) in the north Bay of Bengal has been computed by profile method for the period 18th–25th August 1990 using meteorological data of MONTBLEX-90 from ORVSagarkanya. The fluxes showed synoptic and diurnal variations which are marked during depression (20th–21st August) compared to their variation prior to and after this period. Variations of heat and water vapour fluxes were in phase. Night time fluxes are relatively high compared to day time. Average momentum transfer during depression was two to three times large. Variations in Bowen ratio were relatively large during day time. During depression, it varied between 0·2 in day time and about 0·3 at night and in the undisturbed period between ?0·1 and 0·2 during day time and 0·2 and 0·25 at night. The study shows that the assumptionC _D=C_H=C_E of the exchange coefficients normally used in estimating the fluxes by the bulk aerodynamic method is not appropriate becauseC _H/C_D≈2,C _E/C_D≈1·5 andC _H/C_E≈1·4.     

IR photometry and models of dust shells for two RCB stars

We present JHKLM photometry obtained in 1984–2009 for the RCB stars UV Cas and SU Tau. No major fadings characteristic of RCB stars were detected during the observations of UV Cas, while two events of this kind occurred for SU Tau. The observed flux and color-index variations can be explained with a changing dust concentration in the line of sight, and possibly variations of the stellar temperature. We use the measured fluxes, supplemented with observations in the intermediate IR, to compute spherically symmetric dust-shell models for the stars. The mass-loss rate is estimated to be 1.7 × 10⁻⁶ M _⊙ yr⁻¹ for UV Cas and 4.1 × 10⁻⁶ M _⊙ yr⁻¹ for SU Tau.     

Transport properties of low-sanidine single-crystals, glasses and melts at high temperature

Thermal diffusivity (D) was measured using laser-flash analysis from oriented single-crystal low-sanidine (K_0.92Na_0.08Al_0.99Fe³⁺ _0.005Si_2.95O₈), and three glasses near KAlSi₃O₈. Viscosity measurements of the three supercooled liquids, in the range 10^6.8 to 10^12.3 Pa s, confirm near-Arrhenian behavior, varying subtly with composition. For crystal and glass, D decreases with T, approaching a constant near 1,000 K: D _sat ∼ 0.65 ± 0.3 mm² s⁻¹ for bulk crystal and ∼0.53 ± 0.03 mm² s⁻¹ for the glass. A rapid decrease near 1,400 K is consistent with crossing the glass transition. Melt behavior is approximated by D = 0.475 ± 0.01 mm² s⁻¹. Thermal conductivity (k _lat) of glass, calculated using previous heat capacity (C _P) and new density data, increases with T because C _P strongly increases with T. For melt, k _lat reaches a plateau near 1.45 W m⁻¹ K⁻¹, and is always below k _lat of the crystal. Melting of potassium feldspars impedes heat transport, providing positive thermal feedback that may promote further melting in continental crust.     

A study of turbulent characteristics of atmospheric boundary layer over monsoon trough region using Kytoon and Doppler sodar

As a part of the MONTBLEX-90 observational programme, Kytoon and Doppler sodar observations were taken at Kharagpur. These data are analysed to study the turbulent characteristics of the atmospheric boundary layer in terms of stability, temperature structure function (C _T ² ) and velocity structure function (C _v ² ).C _T ² follows aZ ^−4/3 law on most of the days, whereas the variation ofC _V ² is not systematic.C _V ² andC _T ² values are found to vary between 10⁻⁵−10⁻¹ m^4/3s⁻² and 10⁻⁵−10⁻²°C² m^−2/3 respectively.     

The elastic constants of monoclinic single-crystal chrome-diopside to 1,300 K

Values of the complete adiabatic elastic tensor for single-crystal chrome-diopside (a monoclinic pyroxene mineral) are presented from 298 to 1,300 K. The data were obtained using resonant ultrasound spectroscopy (RUS). They are the first published results for the temperature T dependences of the 13 individual elastic constants C _ij of any clinopyroxene mineral. Each C _ij is appropriately described by a linear function in T throughout the range of T. Values for each (∂C _ij /∂T) _P in GPa K⁻¹ are as follows: C ₁₁, −0.0291; C ₂₂, −0.0248; C ₃₃, −0.0179; C ₄₄, −0.0103; C ₅₅, −0.0077; C ₆₆, −0.0152; C ₁₂, −0.0119; C ₁₃, −0.0064; C ₂₃, 0.0000; C ₁₅, 0.0025; C ₂₅, 0.0022; C ₃₅, −0.0046; and C ₄₆, 0.0026. Values of (∂M/∂T) _P in GPa K⁻¹, where M represents an isotropic bulk property calculated from the C _ij data, are as follows: adiabatic bulk modulus K _S , −0.0123; isothermal bulk modulus K _T , −0.0178; and shear modulus G, −0.00998. Some diopside derivatives, notably (∂K _S /∂T) _P , (∂K _T /∂T) _P , and (∂V _P /∂T) _P , where V _P is the compressional wave velocity, have smaller magnitudes than all other minerals of importance in Earth’s mantle, thus, confirming predictions from systematics studies. We find several dimensionless quantities for this monoclinic mineral have normal values compared to other mantle minerals. Further, αK _T (α is the volume coefficient of thermal expansion) for diopside is approximately independent of both T and volume V at elevated temperature, so its equation of state is accurately expressed in simplified form.     

The use of bond dissociation energy to analyze the geochemical behavior of uranium in the process of granitic magmatism

It is estimated that the uranium-oxygen bond dissociation energy (D _U 4 ₋₀ ⁺ =1465.38kJ/mol) is higher than silicon-and aluminium-oxygen bond dissociation energies (D_Si4 ₋₀ ⁺ =1264.41 kJ/mol and D_Al3 ₋₀ ⁺ =1105.32 kJ/mol). During the process of magmatic differentiation with increasing degree of polymerization for silicon-oxygen complex ion, uranium in granitic melts tends to combine with oxygen to form the coordination polyhedron [UO _x ^2x−n ] and to occur in the form of nuclei and crystals of uraninite, as has been demonstrated by the induced fission-track study of quartz syenite from Huangmeijian.     

Point defects and cation tracer diffusion in (Ti<Subscript><Emphasis Type="Italic">x</Emphasis></Subscript> Fe<Subscript><Emphasis Type="Italic">1−x</Emphasis></Subscript>)<Subscript>3−δ</Subscript>O<Subscript>4</Subscript>. II. Cation tracer diffusion

 Cation tracer diffusion coefficients, D_Me ^*, for Me=Fe, Mn, Co and Ti, were measured using radioactive isotopes in the spinel solid solution (Ti _x Fe _1−x )_3−δO₄ as a function of the oxygen activity. Experiments were performed at different cationic compositions (x=0, 0.1, 0.2 and 0.3) at 1100, 1200, 1300 and 1400 °C. The oxygen activity dependence of all data for D_Me ^* at constant temperature and cationic composition can be described by equations of the type D_Me ^*=D^∘ _Me[V]. C_V·a _O2 ^2/3+D_Me[I] ^∘·a _O2 ^−2/3·D_Me[V] ^∘ and D_Me[I] ^∘ are constants and C_V is a factor of the order of unity which decreases with increasing δ. All log D_Me ^* vs. loga _O2 curves obtained for different values of x and for different temperatures go through a minimum due to a change in the type of point defects dominating the cation diffusion with oxygen activity. Cation vacancies prevail for the cation diffusion at high oxygen activities while cation interstitials become dominant at low oxygen activities. At constant values of x, D_Me[V] ^∘ decreases with increasing temperature while D_Me[I] ^∘ increases.     

Thermal equations of state of dioctahedral micas on the join muscovite–paragonite

 Powder diffraction measurements at simultaneous high pressure and temperature on samples of 2M1 polytype of muscovite (Ms) and paragonite (Pg) were performed at the beamline ID30 of ESRF (Grenoble), using the Paris-Edinburgh cell. The bulk moduli of Ms, calculated from the least-squares fitting of V–P data on each isotherm using a second-order Birch–Murnaghan EoS, were: 57.0(6), 55.1(7), 51.1(7) and 48.9(5) GPa on the isotherms at 298, 573, 723 and 873 K, respectively. The value of (∂K _T /∂T) was −0.0146(2) GPa K⁻¹. The thermal expansion coefficient α varied from 35.7(3) × 10⁻⁶ K⁻¹ at P ambient to 20.1(3) × 10⁻⁶ K⁻¹ at P = 4 GPa [(∂α/∂P) _T = −3.9(1) × 10⁻⁶ GPa⁻¹ K⁻¹]. The corresponding values for Pg on the isotherms at 298, 723 and 823 K were: bulk moduli 59.9(5), 55.7(6) and 53.8(7) GPa, (∂K _T /∂T) −0.0109(1) GPa K⁻¹. The thermal expansion coefficient α varied from 44.1(2) × 10⁻⁶ K⁻¹ at P ambient to 32.5(2) × 10⁻⁶ K⁻¹ at P = 4 GPa [(∂α/∂P) _T = −2.9(1) × 10⁻⁶ GPa⁻¹ K⁻¹]. Thermoelastic coefficients showed that Pg is stiffer than Ms; Ms softens more rapidly than Pg upon heating; thermal expansion is greater and its variation with pressure is smaller in Pg than in Ms. Received: 28 January 2002 / Accepted: 5 April 2002     

<Superscript>210</Superscript>Pb-derived Sedimentation Rates and C<Subscript>org</Subscript> Fluxes in Soledad Lagoon (Cispatá Lagoon System,NW Caribbean Coast of Colombia)

With the aim of evaluating temporal changes in sedimentation and organic carbon (C_org) supplied over the last ~100 years, a sediment core was collected at Soledad Lagoon, a costal ecosystem surrounded by mangroves, located in the Cispatá Estuary (Caribbean coast of Colombia). The core sediments were characterized by low concentrations of calcium carbonate (0.2–2.9%), organic matter (3–8%), total nitrogen (0.11–0.38%), and total phosphorus (0.19–0.65 mg g⁻¹). Fe and Al concentrations ranged from 4% to 5%, and Mn from 356 to 1,047 μg g⁻¹. The ²¹⁰Pb-derived sediment and mass accumulation rates were 1.54 ± 0.18 mm year⁻¹ and 0.08 ± 0.01 g cm⁻² year⁻¹, respectively. The sediment core did not provide evidence of human impact, such as enhancement of primary production or nutrient enrichment, which may result from recent land uses changes or climate change. The C_org fluxes estimated for Soledad Lagoon core lay in the higher side of carbon fluxes to coastal ecosystems (314–409 g m⁻² year⁻¹) and the relatively high C_org preservation observed (~45%) indicate that these lagoon sediments has been a net and efficient sink of C_org during the last century, which corroborate the importance of mangrove areas as important sites for carbon burial and therefore, long-term sequestration of C_org.     

Elasticity of TiO2 rutile to 1800 K

The ambient pressure elastic properties of single-crystal TiO₂ rutile are reported from room temperature (RT) to 1800 K, extending by more than 1200 oK the maximum temperature for which rutile elasticity data are available. The magnitudes of the temperature derivatives decrease with increasing temperature for five of the six adiabatic elastic moduli (C _ij ). At RT, we find (units, GPa): C ₁₁=268(1); C ₃₃=484(2); C ₄₄=123.8(2); C ₆₆=190.2(5); C ₂₃=147(1); and C ₁₂=175(1). The temperature derivatives (units, GPa K⁻¹) at RT are: (∂C ₁₁/∂T) _P =−0.042(5); (∂C ₃₃/∂T) _P =−0.087(6); (∂C ₄₄/∂T) _P =−0.0187(2); (∂C ₆₆/∂T) _P =−0.067(2); (∂C ₂₃/∂T) _P =−0.025; and (∂C ₁₂/∂T) _P −0.048(5). The values for K _S (adiabatic bulk modulus) and μ (isotropic shear modulus) and their temperature derivatives are K _S =212(1) GPa; μ=113(1) GPa; (∂K _S /∂T) _P =−0.040(4) GPa K⁻¹; and (∂μ/∂T) _P =−0.018(1) GPa K⁻¹. We calculate several dimensionless parameters over a large temperature range using our new data. The unusually high values for the Anderson-Gròneisen parameters at room temperature decrease with increasing temperature. At high T, however, these parameters are still well above those for most other oxides. We also find that for TiO₂, anharmonicity, as evidenced by a non-zero value of [∂ln (K _T )/∂lnV] _T , is insignificant at high T, implying that for the TiO₂ analogue of stishovite, thermal pressure is independent of volume (or pressure). Systematic relations indicate that ∂² K _S /∂T∂P is as high as 7×10⁻⁴ K⁻¹ for rutile, whereas ∂²μ/∂T∂P is an order of magnitude less. Received: 19 September 1997 / Revised, accepted: 27 February 1998     

Seasonal variability of deep ocean particle fluxes and particle composition in the north open sea of Prydz Bay

Time-series Mark VII sediment trap was deployed at 72°58.55′E, 62°28.63°S (north of the Prydz Bay, Antarctica) during the cruise of CHINARE-15 in cooperation with University of Marine of America. Seasonal variability of deep ocean particle fluxes and biogenic components were investigated in order to reveal the fluxes and biogeochemistry of sinking particles in the deep ocean. The results show that the total mass flux of sinking particles at a water depth of 1000 m ranges from 13.00 to 334.59 mg⋅d⁻¹⋅m⁻²). A marked seasonal variability exists in the fluxes of all particle components reflecting the seasonal changes in upper water productivity. Biogenic material was a significant component and biogenic silica represented more than 80% of the biogenic matter, reflecting a diatom dominated system, but a lithogenic fraction is always present. The fact that the POC dominated over particulate inorganic carbon (as CaCO₃) and C_inorg/C_org was always greater than 1, indicate a net removal of CO₂ from surface water by biological activity. __________ Translated from Acta Oceanologica Sinica, 2006, 5: 49–55 [译自:海洋 学报]     

Hydrogen partitioning between melt,clinopyroxene, and garnet at 3 GPa in a hydrous MORB with 6 wt.% H<Subscript>2</Subscript>O

To understand partitioning of hydrogen between hydrous basaltic and andesitic liquids and coexisting clinopyroxene and garnet, experiments using a mid-ocean ridge basalt (MORB) + 6 wt.% H₂O were conducted at 3 GPa and 1,150–1,325°C. These included both isothermal and controlled cooling rate crystallization experiments, as crystals from the former were too small for ion microprobe (SIMS) analyses. Three runs at lower bulk water content are also reported. H₂O was measured in minerals by SIMS and in glasses by SIMS, Fourier Transform infrared spectroscopy (FTIR), and from oxide totals of electron microprobe (EMP) analyses. At 3 GPa, the liquidus for MORB with 6 wt.% H₂O is between 1,300 and 1,325°C. In the temperature interval investigated, the melt proportion varies from 100 to 45% and the modes of garnet and clinopyroxene are nearly equal. Liquid composition varies from basaltic to andesitic. The crystallization experiments starting from above the liquidus failed to nucleate garnets, but those starting from below the liquidus crystallized both garnet and clinopyroxene. SIMS analyses of glasses with >7 wt.% H₂O yield spuriously low concentrations, perhaps owing to hydrogen degassing in the ultra-high vacuum of the ion microprobe sample chamber. FTIR and EMP analyses show that the glasses have 3.4 to 11.9 wt.% water, whilst SIMS analyses indicate that clinopyroxenes have 1,340–2,330 ppm and garnets have 98–209 ppm H₂O. D _H ^cpx−gt is 11 ± 3, D _H ^cpx−melt is 0.023 ± 0.005 and D _H ^gt−melt is 0.0018 ± 0.0006. Most garnet/melt pairs have low values of D _H ^gt−melt, but D _H ^gt−melt increases with TiO₂ in the garnet. As also found by previous studies, values of D _H ^cpx−melt increase with Al₂O₃ of the crystal. For garnet pyroxenite, estimated values of D _H ^{pyroxenite−melt} decrease from 0.015 at 2.5 GPa to 0.0089 at 5 GPa. Hydration will increase the depth interval between pyroxenite and peridotite solidi for mantle upwelling beneath ridges or oceanic islands. This is partly because the greater pyroxene/olivine ratio in pyroxenite will tend to enhance the H₂O concentration of pyroxenite, assuming that neighboring pyroxenite and peridotite bodies have similar H₂O in their pyroxenes. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.     

Sensitivity of high-resolution tropical cyclone intensity forecasts to surface flux parameterization

Surface flux parameterization schemes used in current dynamic models are primarily based upon measurements at low and moderate wind speeds. Recent studies show that these parameterization schemes may be incorrect at high wind speeds (e.g., tropical cyclone forecasts). Five high-resolution numerical model experiments are designed to assess the sensitivity of tropical cyclone intensity forecasts to changes in the surface flux parameterization. The sensitivity experiments are conducted by running 48 h forecasts of the Coupled Ocean/Atmosphere Mesoscale Prediction System (COAMPS) for six selected tropical cyclones with individual modifications to surface flux calculation that include: (1) limiting the surface stress for wind speeds greater than 33 m s⁻¹, or 64 knots (kt); (2) computing the stress at the top of the model bottom grid layer (MBGL) by averaging results from surface layer similarity and turbulence mixing parameterization for wind speeds greater than 33 m s⁻¹; (3) increasing the roughness lengths for heat and moisture transfer by a factor of ten; (4) setting the roughness lengths for heat and moisture transfer to 1/10 of the momentum roughness length; and (5) cooling the sea surface temperature (SST) by a prescribed rate at high winds. Averaged responses for the six storms to these sensitivity tests show that: (i) the limit on surface stress at high winds significantly increases the cyclone intensity in 48 h forecasts; (ii) the averaged surface layer stress at high winds increases the cyclone intensity but to a much lesser degree than limiting the surface stress; (iii) large increases in the roughness lengths for heat and moisture transfer are needed to significantly impact the intensity forecast; (iv) the different roughness length formula for surface transfer coefficients notably increases C _h/C _d ratio from 0.59 to 0.79 for 25 m s⁻¹ and 0.41 to 0.75 for 50 m s⁻¹ that significantly increases the predicted cyclone intensity; and (v) cooling of the SST by −5.8°C in 48 h reduces the maximum surface wind speed by −32 kt, or 16.5 m s⁻¹, at 48 h forecast. These results suggest that a surface flux parameterization scheme suitable for tropical cyclone intensity forecast must correctly model the leveling-off character of surface stress and C _h/C _d ratio at high winds. All modifications to surface flux calculation have little influence on 48 h track forecasts, even though they may significantly impact the intensity forecasts.

The Partitioning of Tungsten bwtween Aqueous Fluids and Silicate Melts

An experimental study has been carried out to determine the partition coefficients of tungsten between aqueous fluids and granitic melts at 800 °C and 1.5 kb with natural granite as the starting material. The effects of the solutions on the partition coefficients of tungsten show a sequence of P > CO ₃ ²⁻ > B > H₂O. The effects are limited (generallyK _D < 0.3) and the tungsten shows a preferential trend toward the melt over the aqueous fluid. The value ofK _D increases with increasing concentration of phosphorus; theK _D increases first and then reduces with the concentration of CO ₃ ²⁻ when temperature decreases, theK _D between the solution of CO ₃ ²⁻ and the silicate melt increases, and that between the solution of B₄O ₇ ²⁻ and the silicate melt decreases. The partition coefficients of phosphorus and sodium between fluids and silicate melts have been calculated from the concentrations of the elements in the melts. TheK _D value for phosphorus is 0.38 and that for sodium is 0.56. Evidence shows that the elements tend to become richer and richer in the melts.     

Trace element partitioning between orthopyroxene and anhydrous silicate melt on the lherzolite solidus from 1.1 to 3.2 GPa and 1,230 to 1,535°C in the model system Na<Subscript>2</Subscript>O–CaO–MgO–Al<Subscript>2</Subscript>O<Subscript>3</Subscript>–SiO<Subscript>2</Subscript>

We determined experimentally the Nernst distribution coefficient between orthopyroxene and anhydrous silicate melt for trace elements i in the system Na₂O–CaO–MgO–Al₂O₃–SiO₂ (NCMAS) along the dry model lherzolite solidus from 1.1 GPa/1,230°C up to 3.2 GPa/1,535°C in a piston cylinder apparatus. Major and trace element composition of melt and orthopyroxene were determined with a combination of electron microprobe and ion probe analyses. We provide partitioning data for trace elements Li, Be, B, K, Sc, Ti, V, Cr, Co, Ni, Rb, Sr, Y, Zr, Nb, Cs, Ba, La, Ce, Sm, Nd, Yb, Lu, Hf, Ta, Pb, U, and Th. The melts were chosen to be boninitic at 1.1 and 2.0 GPa, picritic at 2.3 GPa and komatiitic at 2.7 and 3.2 GPa. Orthopyroxene is Tschermakitic with 8 mol% Mg-Tschermaks MgAl[AlSiO₆] at 1.1 GPa while at higher pressure it has 18–20 mol%. The rare earth elements show a continuous, significant increase in compatibility with decreasing ionic radius from D _La^opx−melt ∼ 0.0008 to D _Lu^opx−melt ∼ 0.15. For the high-field-strength elements compatibility increases from D _Th^opx−melt ∼ 0.001 through D _Nb^opx−melt ∼ 0.0015, D _U^opx−melt ∼ 0.002, D _Ta^opx−melt ∼ 0.005, D _Zr^opx−melt ∼ 0.02 and D _Hf^opx−melt ∼ 0.04 to D _Ti^opx−melt ∼ 0.14. From mathematical and graphical fits we determined best-fit values for D ₀^M1, D ₀^M2, r ₀^M1, r ₀^M2, E ₀^M1, and E ₀^M2 for the two different M sites in orthopyroxene according to the lattice strain model and calculated the intracrystalline distribution between M1 and M2. Our data indicate extreme intracrystalline fractionation for most elements in orthopyroxene; for the divalent cations D _i ^M2−M1 varies by three orders of magnitude between D _Co^M2−M1 = 0.00098–0.00919 and D _Ba^M2−M1 = 2.3–28. Trivalent cations Al and Cr almost exclusively substitute on M1 while the other trivalent cations substitute on M2; D _La^M2−M1 reaches extreme values between 6.5 × 10⁷ and 1.4 × 10¹⁶. Tetravalent cations Ti, Hf, and Zr almost exclusively substitute on M1 while U and Th exclusively substitute on M2. Our new comprehensive data set can be used for polybaric-polythermal melting models along the Earth’s mantle solidus. Electronic supplementary material  The online version of this article (doi:) contains supplementary material, which is available to authorized users.