EXPLANATION OF DIAGNOSTICS *****BUDGETS***** Values are presented of various quantities as a function of latitude as well as hemispheric and global integrals for different surface types: global, open ocean, ocean ice, (total) ocean, land, land ice, open lakes, lake ice, (total) lakes. And then the same quantities are provided for specific regions (East U.S., Amazon Rainforest, etc.). Many of the quantities/terminology are obvious; discussed below are those which are not. P0 refers to the top of the model; P1 refers to the dynamic top. Above the level where atmospheric dynamics is calculated are three atmospheric radiation levels, representing primarily the thermosphere, to absorb high frequency energy that should not get down to the lower, computational levels. These then are above P1, while P0 is the top of the highest of them. Z0 refers to the top of the ground. SW is shortwave; LW is longwave; ALB is albedo; ABS is absorbed; LW WINDOW BTEMP is the longwave radiation at 10 microns at P0; NET CLR RAD is the net clear-sky radiation; NET CLR RAD TRP; NET RAD (TROPP) are the net clear sky radiation and the net radiation at the tropopause; HT RVR DISCH is the discharge of heat by rivers; HEAT RUNOFF is the heat removed by river runoff; SS Precip is the precipitation calculated in the large-scale supersaturation subroutine; MC Precip is the precipitation calculated in the moist convection subroutine; H2O BY CH4 is the parameterized generation of water vapor by methane oxidation in the upper stratosphere; IRR ADD is the irrigation water added from an external source, or from ground water. IRRE AD is the heat added from this external source. TG1 and TG2 are the first two layers of the ground; T Surf is the surface air temperature, T Air is the vertically integrated atmospheric temperature, and T1 is the temperature of the first layer of the atmosphere (these are all shown in tenths of degrees); DT/DLAT is the temperature change with latitude; STAT STB is the static stability, given for both the troposphere and stratosphere, from the vertical temperature gradient between the top and bottom of each region; RICH Num is the Richardson Number, a measure of local vertical temperature instability; Ross Num is the Rossby Number, a ratio of the nonlinear term to the coriolis force in the equation of motion; when order of one or less (large-scale) flows are in geostrophic balance, between the pressure gradient and coriolis forces; L is the Rossby radius of deformation, the length scale at which rotational effects become as important as buoyancy or gravity wave effects in the evolution of the flow. At mid-latitudes in the troposphere it is on the order of 1500-2000 km; length scales larger than that have rotation as their predominant restoring force (e.g., Rossby waves), length scales smaller have buoyancy/gravity as the restoring force (e.g., gravity waves). Directly proportional to buoyancy, and inversely proportional to the coriolis force, these scales will be very different on different planets (e.g., Venus). GAM is the lapse rate; GAMM is the moist adiabatic lapse rate (lapse rates higher than that are convectively unstable); GAMC is the baroclinic instability lapse rate, values higher than that are unstable dynamically so produce wave/eddy kinetic energy; MC CLD DPTH is the moist convective cloud depth; OCEAN TRNS CONV is the heat transport convergence by ocean dynamics (in a Q-flux model, the value specified); SURF TYPE FRACT relates to the surface type (land, ocean, lakes, etc.). *****ZONAL AVERAGE LATITUDE X ALTITUDE***** There are two coordinate systems for the latitude x altitude diagnostics Ð constant pressure diagnostics, and constant sigma diagnostics sigma=(P-Ptop)/(Ps-Ptop) (=1 at the surface, =0 at P=Ptop) where Ptop is the top of that part of the atmosphere in which we use sigma coordinates rather than constant pressure coordinates. For convenience sake, sigma may be extended above Ptop by replacing in that formula the local Ps by the global_mean_Ps, and that sigma then is negative in that domain. In the order of the print out, the first set of diagnostics (through Dtemp/Dt by Dynamics) are all constant pressure diagnostics Ð they often have (CP) next to the title. The following ones (beginning with ÔVertical mass exchange by moist convectionÕ) are then all constant sigma diagnostics Ð with one major caveat. In all our current models, we change over running the model in constant sigma to running in constant pressure coordinates at 150 mb; therefore, even the sigma coordinate diagnostics are really constant pressure at altitudes above 150 mb. Below 150 mb, for the CP diagnostics, they are interpolated to constant pressure, keeping the vertical integral constant. The diagnostics as shown in the print file are at a coarse resolution (8¡ x 10¡), but they are actually saved at a finer resolution, which can be printed out if needed. NUMBER OF GRIDPOINTS INCLUDED IN AVERAGE: Because of mountains, some longitudes do not have an atmosphere at the higher pressure levels near the surface. This diagnostics tells you how many longitudes are actually contributing to the zonal average. Above about 565 mb, all the longitudes (144 at 2x2.5 resolution) contribute. PRESSURE DIFFERENCES: This diagnostic gives the difference in mean pressure between subsequent layers. As noted in the title, it is on the UV grid. With the B grid dynamics schemes, pressure, height, temperature and humidity are calculated at the center of a grid box, while the U and V winds are calculated at the grid box edges. So the ÔUVÕ grid is offset from the standard Temp, etc. grid by 2¡ in latitude. If the highest northern latitude is 90¡, the diagnostic is on the Temp grid. If it is 88¡, it is on the U grid. STANDARD DEVIATION OF PRESSURE DIFFERENCES: Again because of mountains, not all layers are present at low altitudes, and hence the pressure differences between layers are not the same at each longitude along a latitude circle. This diagnostic indicates the standard deviation of the pressure difference values due to this effect among others at each latitude. Above about 565 mb, the values are generally zero. TEMPERATURE (T): The value is given for each latitude x pressure, with the vertically integrated value at the bottom multiplied by 0.1 for greater resolution. This is often an interesting number to show in climate change experiments, for while the surface temperature may indicate greater warming at high latitudes, the vertically-integrated value often does not. The three highest levels are the so-called Ôradiative layersÕ, in which atmospheric radiation calculations are carried out, but no dynamics Ð there is a rigid lid at their base, although this does not prevent radiative interactions with the layers below. These layers are here to block the extreme UV radiation, etc. which is absorbed in the thermosphere from entering the lower atmospheric levels. HEIGHT: Literally Ôgeopotential heightÕ, it is similar to geometric height except for the slight variation of gravity with altitude. It too shows the values of the three radiative layers at the top. SPECIFIC HUMIDITY (q): the ratio of the total number of water vapor molecules in a unit volume of air to the total number of molecules in a mixture of the dry air + water vapor, when expressed in ppmv (as done here). It can also be given in terms of the mass of water vapor/mass of the mixture of dry air + water vapor, in which case it is ppm. If given in those units, the values are lower by the ratio of water vapor mass to dry air mass, or 18/28 (0.643). RELATIVE HUMIDITY: the ratio of the partial pressure of water vapor in an air-water mixture to the saturated vapor pressure of water vapor at a prescribed temperature Ð in other words, the amount of moisture in the air divided by the maximum amount of moisture the air can hold at that temperature (when saturated). Measured in percent. When the predominant particles are ice, as determined by the model, the calculation is done with respect to ice (and as the saturation vapor pressure with respect to ice is less than for water, the relative humidity would be greater over ice than over water at the same temperature). TOTAL CLOUD WATER CONTENT, CLOUD LIQUID WATER CONTENT, CLOUD ICE WATER CONTENT: The calculated condensate in clouds in the model, with the ice and liquid water content adding up to the total. Ice is calculated below a threshold temperature or with a probability distribution, depending on the scheme used (greater probability of ice, the lower the temperature below 0¡C). These values will all change with climate, for example, in the upper troposphere there will be a decrease in cloud ice and increase in cloud water content. As condensation is easier over ice (lower saturation vapor pressure required), that may reduce high level cloud cover in some regions. DT(MC)*P DRY HEATING; DT(MC)*P CHANGE OF PHASE; CHANGE IN TOTAL WATER BY MOIST CONV: (newer diagnostics so their location may vary): A breakdown of the different moist convective process contributions to energy, including from altering the water content in the atmosphere. ZONAL WIND (U), MERIDIONAL WIND (V): The zonal wind is positive from west to east, and the meridional wind positive from south to north. The values around 200 mb at around 30¡ latitude generally constitute the time-averaged jet stream numbers. V* = V - D(V'TH'/DTHDP)/DP : One of the so-called Ôtransformed EulerianÕ (TEL) diagnostics, in this case the TEL meridional velocity. The idea is that a zonal average meridional wind can be generated by zonal average heating (e.g., warm air in the tropics rising and moving poleward). It can also be generated by the warmth created by wave-induced convergences (an eddy or storm system bringing warm air into a region). So even though these are zonal average diagnostics, an eddy (defined as a deviation at each longitude from the zonal average) could be influencing the results by setting up its own zonal average circulation, which in the extratropics can overwhelm the one caused by zonal average heating. The TEL diagnostics try to remove the eddy contribution, and leave just the zonal average forcing, by subtracting out the eddy induced warming effect (vÕ?Õ divided by a measure of stability, d?/dP, where ? is potential temperature). Comparison to V just above it shows that the two diagnostics are fairly similar in the tropics, where eddies are weak, and diverge considerably in the extratropics. Without the contribution from eddies, the annual average V* shows flow from the tropics to the poles in the upper troposphere, with return flow below, while V shows a three cell distribution (Hadley Cell, Ferrel Cell, Polar Cell), with the flow from the tropics ending at around 30¡N, S. As climate and the heating distribution changes, this will indicate how the poleward flow induced by that heating changes, without the interfering effects of eddy changes. TEL diagnostics can better directly represent the net movement of tracers. Since both versions ultimately must show the same result, in the TEL framework, tracers are transported directly from low latitudes to high latitudes in the upper troposphere, while in the standard Eulerian approach, the mean circulation only brings them to 30¡N,S, while eddies pick them up and transport them poleward from there. STREAM FUNCTION: Really, the Ômass streamfunctionÕ, generated by the variation of meridional wind with altitude and vertical wind with latitude. The convention is that a negative value indicates a counterclockwise circulation in the frame of the printout (NH on the left), and a positive value a clockwise circulation. On the annual average, or in winter, you can thus see a Ôthree cellÕ distribution in both hemispheres. A prime diagnostic for indicating how the Hadley Cell will change with climate. TRANSFORMED STREAM FUNCTION: Utilizing V* as calculated above (and a corresponding W*, since eddy convergences of heat also generate a vertical motion), this is the TEL equivalent of the standard mass streamfunction. On the annual average you thus see a direct circulation from equator to pole in each hemisphere. Again, it represents the movement of tracers. W* RESIDUAL VERTICAL VELOCITY: The TEL vertical velocity, also called (as here) the residual vertical velocity, since it is what is left after subtracting out the eddy component. [Hence the TEL circulation can also be called the residual circulation]. VERTICAL VELOCITY: In pressure coordinates, it is literally dp/dt, and hence normally it is positive downward, but for the sake of comparison with W*, what is printed out is ÐW (i.e., positive upward). Comparison with W* just above shows substantial differences in the extratropics, less in the tropics where eddies are weak. STANDING EDDY KINETIC ENERGY, EDDY KINETIC ENERGY, TOTAL KINETIC ENERGY: The U or V wind at any location can be defined as the sum of the latitudinal average value plus a deviation from the zonal average necessary to produce the observed value. The deviation, or eddy, is generally the result of wave-like phenomena. These waves can move, associated for example with storm systems, or be relatively fixed in space, associated for example with the land/ocean contrast or topography. The kinetic energy (per unit density) is calculated as the sum of the squares of the U and V components; the Standing Eddy Kinetic Energy is the value calculated from these ÔdeviationsÕ averaged over a month. The Eddy Kinetic Energy is the sum of both the standing and traveling components. The Total Kinetic Energy is the sum of the Eddy Kinetic Energy and the zonal or latitudinal average value, i.e., ÔZonal Kinetic EnergyÕ, again calculated as the sum of the squares of the latitudinal average U and V components. To obtain the Zonal Kinetic Energy subtract the Eddy Kinetic Energy from the Total Kinetic Energy. Standing eddy kinetic energy should be much less on an aquaplanet. POTENTIAL TEMPERATURE (theta): As air rises it cools due to the decrease in density, so its temperature drops even though there is no loss of ÔheatÕ from the parcel. To compensate for that, the potential temperature is calculated by conceptually bringing the air downward without gain or loss of heat to 1000 mb elevation (it thus warms at 10¡C/km). [Potential temperature is actually the variable that the model calculates, rather than temperature]. When air is well-mixed, potential temperature is relatively uniform with altitude, although under time-averaged conditions, it increases somewhat with altitude everywhere, indicating stability. Because air is not well-mixed between the troposphere and stratosphere, there is a big jump in potential temperature between these levels. POTENTIAL VORTICITY: Relative vorticity by itself it the curl of the wind vector, in the horizontal plane that mean [dv/dy Ð du/dx]. As such, it involves only the equations for conservation of momentum. Absolute vorticity is the sum of the coriolis parameter f and the relative vorticity. Potential vorticity, in this form, brings in the energy equation as well by multiplying the absolute vorticity by a representation of stability {d?/dp}. In that sense, potential vorticity is a more powerful variable. As in the case of potential temperature, it too can be used to indicate the transition from the (dynamic) troposphere to the stratosphere. To produce what are called ÔPVÕ units (or PVU), the values shown here have to be divided by 10 (switching to units of hectopascals and then multiplying by gravity); the transition from the troposphere to the stratosphere is then thought to occur in the region of 1-3 PVU. As climate changes, this transition region may rise. NORTHWARD TRANSPORTS OF SENSIBLE HEAT; DRY STATIC ENERGY; LATENT HEAT; STATIC ENERGY; KINETIC ENERGY: The different forms of energy in the atmosphere are sensible heat (temperature, T), geopotential energy (geopotential height, phi), latent heat (specific humidity, q) and kinetic energy (U2 + V2). The kinetic energy transport is small relative to the other three. Northward transport of sensible heat is calculated as the meridional velocity V x T; northward transport of geopotential energy V x phi; northward transport of latent heat, V x q. Dry static energy is the sum of sensible + geopotential energy. Static energy (sometimes referred to as Ômoist static energyÕ) is the sum of dry static energy + latent heat. To obtain the geopotential energy transport from these diagnostics, subtract the sensible heat transport from the dry static energy transport; one will find there is little difference in the extratropics. However in the tropics there is a big difference; there convection and vertical motion lift the air upward at low latitudes (increasing its geopotential energy), then move it poleward to ~30¡N,S, where it descends. The northward transport of energy by the low latitude Hadley Cell is primarily in the form of geopotential energy. As climate warms, and the latitudinal temperature gradient decreases, transport of sensible heat and dry static energy to the pole should decrease. But with more water vapor in the air, latent heat transport increases. So the transport of static energy to high latitudes may or may not decrease. A note on the ÔhemisphericÕ values of the northward transports indicated on the diagnostics. These mass-weighted transports indicate the effective ÔpolewardÕ movement of species, and one can compare that tendency from one model or climate simulation to another. Since the transport is occurring at different latitudes in different ways, it does not have any specific meaning other than that. NORTHWARD TRANSPORTS BY STANDING EDDIES; EDDIES; AND TOTAL TRANSPORT. The above quantities are broken down into the components of transports by the standing, and then standing + traveling eddies [ÔeddyÕ transport]; to get just the traveling component, subtract the standing eddy transport from the eddy transport. The total transport also includes transports by the mean circulation; to get its transports alone, subtract the eddy transport from the total transport. As climate warms eddy energy may decrease, and therefore their transports may change for that reason as well. On an aquaplanet, without standing energy, these transports will be lost Ð will they be picked up by the traveling eddies? By the ocean? The transports can be used to diagnose the energy changes at any latitude due to changes in atmospheric dynamics. One would calculate the convergence of the energy Ð how much is leaving the latitude to the north minus what is coming in from the latitude to the south. Often this approach utilizes the vertical integral at the bottom of the respective latitudes as a convenient summary for the latitudinal column as a whole. It could then be compared to the radiative induced changes (shown in output below). [There are two different versions done for latent heat Ð the differences between them are generally small]. Because the kinetic energy transport is small, only its total transports is shown. NORTHWARD TRANSPORT OF ZONAL MOMENTUM BY STANDING EDDIES; EDDIES;AND TOTAL: zonal momentum (per unit mass) is represented by the zonal wind (U), and so its northward transport is V x U. As in the case of the energy transports, it is broken down into a transport contribution from the different atmospheric features. Waves generally transport momentum in the opposite direction from their meridional movement. So the northward transport of momentum by eddies at latitudes south of 50¡N implies the wave energy (planetary waves, not necessarily storms) is propagating equatorward, while the southward transport at higher latitudes implies the waves there are propagating poleward. The momentum convergence at around 50¡N helps generate the polar front. DYNAMIC CONVERGENCE OF DRY STATIC ENERGY; to simply the energy convergence calculation noted above, this diagnostic does it for you (primarily depends on the northward transports, and vertical transports and convergences are small). DYNAMIC CONVERGENCE OF EDDY GEOPOTENTIAL: here the vertical transports are quite important, and so the convergence calculation shows their influence. See the discussion with the next diagnostic for its importance. BAROCLINIC EDDY KINETIC ENERGY GEN.: the primary forcing for waves in the extratropoics (on Earth) comes from the rising of light warm air and sinking of dense cold air, which amounts to a loss of geopotential energy and a gain of kinetic energy. This is the ÔbaroclinicÕ process, literally indicating inclined pressure surfaces relative to density surfaces. This diagnostic (-omegaÕ x alphaÕ, where omegaÕ is the eddy vertical velocity in pressure coordinates, and alphaÕ is the eddy specific volume [=1/densityÕ]) shows where it occurs in the Northern Hemisphere, primarily at mid latitudes in the mid-troposphere, and in the Southern Hemisphere also at higher latitudes, where the contrast between the cold air coming off Antarctica contrasts with the warmer air over the ocean. This too will change with climate, and its change can then be related to the change in eddy kinetic energy. However, even though the generation is greatest in the mid-troposphere, the eddy kinetic energy diagnostic indicates the eddy energy is not largest there. ThatÕs because the energy that is generated is transported both upward and downward from that region, primarily in the form of geopotential energy. Where that energy converges Ð as shown in the Ôdynamic convergence of eddy geopotential diagnosticÕ is where the energy would really peak. P-K BY EDDY PRESSURE GRADIENT FORCE: This diagnostic which literally indicates the transformation of potential to kinetic energy by the pressure gradient force, in effect acts as the sum of where the energy is being generated plus where the transport is having it converge (i.e., the baroclinic eddy kinetic energy generation + the Dynamic convergence of eddy geopotential). It then says that the maximum energy being generated should be in the lower troposphere and upper troposphere. The eddy kinetic energy diagnostic says it is actually in the upper troposphere; the lower tropospheric eddy energy is destroyed by turbulence/friction. VERT. TRANS. OF GEOPOT. ENERGY BY EDDIES: As noted above, while the baroclinic generation is actually greatest in the mid-troposphere, vertical transport of geopotential energy moves it up and down from there. This is the diagnostic that shows that. It provides a basic indication of how wave energy propagates from the troposphere to the stratosphere. Note that in all the vertical transport diagnostics, even though they are shown in pressure coordinates, the transport is calculated in the sense that it is positive upward. Also, with respect to the mass-weighted vertical integrals of the vertical transports: this indicates the effective ÔupwardÕ movement of species, and one can compare that tendency from one model or climate simulation to another. Since the transport is occurring at different levels in different ways, it does not have any specific meaning other than that. VERT. TRANS. OF DRY STATIC ENERGY BY EDDIES: Similar to the northward transports, this now combines the vertical transport of sensible heat + geopotential energy. To get the sensible heat by itself, subtract the vertical transport of geopotential energy by eddies from this diagnostic. TOTAL LGE SCALE VERT. TRANS. OF DRY STAT ENRG: combines the eddy and mean circulation transports of this quantity together. As indicated by the units in this diagnostic relative to those for the eddy vertical transport, it is dominated by the mean circulation effects; the Hadley Cell lifting of energy in the tropics and descent in the subtropics is clearly evident. VERTICAL TRANSPORT OF LATENT HEAT BY EDDIES; LARGE-SCALE VERTICAL TRANSPORT OF LATENT HEAT: transport of latent heat by eddies, and then by the sum of eddies and the mean circulation. To get the mean circulation effects by themselves, subtract the first diagnostic from the second. Both eddies and the mean circulation play a significant role. VERTICAL TRANSPORT OF STATIC ENERGY BY EDDIES; LARGE-SCALE VERTICAL TRANSPORT OF STATIC ENERGY BY EDDIES. The sum of the vertical transport of dry static energy + latent heat, first by eddies, and then by eddies plus the mean circulation. Mean circulation effects dominate. TOTAL LGE SCALE VERT. TRANS. OF KINETIC ENRG. The sum of eddy + mean circulation transports of kinetic energy. Note how small these values are, relative to the vertical transport of the other energy forms. EDDY VERTICAL ZONAL MOMENTUM FLUX; TOTAL VERTICAL ZONAL MOMENTUM FLUX: the vertical transport of zonal momentum (U x vertical velocity, W) first by eddies, and then by eddies plus the mean circulation. These values are small relative to other influences on the zonal wind. VERTICAL TRANSPORT OF POTENTIAL VORTICITY; VERTICAL TRANSPORT OF POTENTIAL VORTICITY BY EDDIES. The vertical transport of potential vorticity clearly shows where air is being transported downward into the troposphere from the stratosphere. Similarly, upward transport shows where it being sent from the troposphere into the stratosphere. {To represent this in PV units, use the scaling factor Ð dividing by 10 Ð given previously for potential vorticity}. The first diagnostic includes both eddy and mean circulation effects, the second the eddy transport alone. Even though the eddy flux is usually the one discussed in this respect, the diagnostics show that the mean circulation influence is not negligible. NORTH. TRANS. OF EDDY Q-G POT. VORTICITY. This is the first of the quasi-geostropohic (Q-G) diagnostics. The Q-G approaches basically linearizes the equations of motion so as to make them analytically more tractable. It works very well in Q-G models that are the theoretical basis for much instability theory. It does not work as well in the fully non-linear models like GCMs, but are included here as most theoretical interpretations of atmospheric dynamics use such diagnostics. In particular, baroclinic instability is associated with equatorward transport of Q-G potential vorticity, shown occurring here in the mid and upper troposphere. Comparison to the Ôbaroclinic eddy kinetic energy generationÕ diagnostic shows a general correspondence, although that diagnostic shows a peak in the middle troposphere, where here the peak is towards the upper troposphere. Q-G POT. VORTICITY CHANGE OVER LATITUDES. Literally d(Q-G)/dy, where the value of this diagnostic changes sign in latitude or altitude, it indicates instability, barotropic instability (eddy energy derived from the zonal mean flow) with a change in sign with latitude, and baroclinic instability (eddy energy derived from the potential energy of the temperature/density field) with a change in sign in the vertical. One can see both aspects occurring in the NH mid-troposphere. REFRACTION INDEX FOR WAVE NUMBERS 1,2,3,6 and 9: This is another Q-G diagnostic. The sense is that where this diagnostic is negative, wave energy is blocked from propagating vertically, especially where the region of negativity has a large vertical extent. One can see that it is more negative at higher wavenumbers, implying those waves have more difficulty propagating into the upper troposphere and stratosphere. It helps explain why wave energy shifts to lower wavenumbers with an increase in altitude. As with other Q-G diagnostics, it gives a sense of whatÕs happening, but is not precise for the GCM. One can also see the great difficulty waves would have propagating vertically in the tropics and high latitudes. DU/DT diagnostics: all of these attempt to explain why the zonal wind has changed during the month. DU/DT total change: this simply looks at the zonal wind at the end of the month minus that at the beginning to see how it really did change. DU/DT BY EULER CIRC. + CONVEC + DRAG+DIF+ER2: This diagnostic, in the standard Eulerian framework, adds all the difference diagnostics that are indicating mechanisms which have changed in the zonal wind. In a perfect world, this diagnostic would equal the total change one. It certainly doesnÕt do that in the lower troposphere, where frictional effects and sigma coordinates interfere, and is better in the stratosphere, which is in constant pressure coordinates. DU/DT BY MEAN ADVECTION: Again in the standard Eulerian framework, this indicates how advection by the mean circulation (via transport of momentum) has changed the zonal wind during the month. DU/DT BY EDDY CONVERGENCE: In the standard Eulerian framework, the convergence of angular momentum by eddies will also change the zonal wind, detailed here. DU/DT BY TRANSFORMED ADVECTION: This is a transformed Eulerian (TEL) diagnostic, which indicates how the mean circulation that is not being generated by eddy convergences has altered the zonal wind (i.e., the streamfunction that is being generated by heating and friction). DU/DT BY ELIASSEN-PALM DIVERGENCE (EPFD): Eliassen-Palm fluxes are fluxes of wave energy normalized by the zonal wind; where they diverge they accelerate west winds, and where they converge, they weaken them. This is often referred to as the total eddy influence on the zonal wind. Why doesnÕt the eddy convergence of momentum, in the standard Eulerian framework, indicate the total eddy influence on the zonal wind? It is because in that framework eddies also transport and converge energy Ð that convergence warms the air and causes it to rise, and move poleward or equatorward, where it is turned by the coriolis force Ð which then also alters the zonal wind. In the TEL framework, the eddy transport of sensible heat has been subtracted out (in the process producing a ÔtransformedÕ circulation), so it no longer contributes. The fact that the EPFD is approximately the total eddy effect is one reason for its popularity. Together, DU/DT by [Transformed Advection + EPFD ]equals DU/DT by [Mean advection + eddy convergence], i.e., the sum of the terms in the standard Eulerian framework have to equal the terms in the transformed Eulerian framework, mathematically. DU/DT by FD error terms, 1 and 2. These are finite difference errors that arise in the calculations. Generally they are small, especially the first one, while the second tends to be largest near the model top. When comparing the sum of all the calculated changes to the Ôtotal changeÕ that actually occurred, sometimes including them improves the comparison. DU/DT by STRAT MTN Drag; by STRAT DEFORM Drag; by STRAT SHEAR Drag; by STRAT MC Drag (+/- 10m/s; +/- 40m/s; +/-20m/s): these are all the changes in the zonal wind produced by the parameterized gravity wave drags associated with different model processes (flow over mountains, deformation, shear and convection). They are there primarily to help produce a better stratosphere and tropopause, and are thought to be operative in the real world. Not all were active in the standard Model E formulation. When momentum is lost in, say, a stratospheric region due to this process, it is put back in the generating region for the tropospheric phenomenon, primarily in the troposphere, so the process conserves momentum. ZONAL WIND CHANGE BY MTN+DEFORM+SHR+MC DRAG: The sum of the contributions of all these parameterized gravity wave drags to the change in zonal wind. DU/DT BY GRAVITY WAVE MOMENTUM DIFFUSION: In previous versions of the model, when the parameterized gravity waves were calculated to ÔbreakÕ and directly damp the mean wind, they were also calculated to diffuse horizontal momentum. This has now been disabled for scaling reasons. DU/DT BY VERTICAL DIFFUSION: When the zonal/meridional wind has gradients that are too sharp in the vertical, it is presumed that inertial instability will reduce the gradients, which this diagnostic does. ItÕs most effective in the tropical upper stratosphere, where observations show the process seems to occur. DU/DT BY SDRAG: A modified ÔRayleigh frictionÕ (or really, modified surface friction) is used to minimize the influence of the model top, and it is called ÔSDragÕ. It has a specified latitude/altitude distribution where it is active. DTEMP/DT TOTAL CHANGE: Similarly, the change in temperature that actually occurred over the course of the month is tabulated from output at the beginning and end of the month. Note this includes both the dynamical changes shown just below, and the radiative-induced changes. DTEMP/DT BY MEAN ADVECTION; DTEMP/DT BY EDDY CONVERGENCE: The two dynamical changes of temperature from the standard Eulerian framework, first by the mean circulation and then by eddies. DTEMP/DT BY TRANSFORMED ADVECTION: The equivalent from the TEL framework. One of the TEL advantages is that there is no eddy contribution to the temperature change (as it has been subtracted out), so the two Eulerian diagnostics for temperature change should equal this one. DTEMP/DT BY STRATOSPHERIC DRAG: As the SDrag is reducing the wind, and hence the kinetic energy, that energy is put back into temperature. DTEMP/DT BY DYNAMICS: The temperature is calculated before and after the dynamics time-step, and so the change induced by the dynamics is recorded. It can be compared with that calculated in the Eulerian or TEL diagnostics. The comparison is not perfect, especially in the lower troposphere, but the calculated results are representative of what the model was actually doing. This ends the constant pressure level diagnostics. As to why the following diagnostics are not in constant pressure coordinates? Partly it was thought that constant pressure was more important for some than for others, and partly no one pushed to have them put in constant pressure. ------------------------------------------- VERTICAL MASS EXCHANGE FROM MOIST CONVECTION: Despite the name, this is an upward flux of mass in the plume; compensatory subsidence and downdrafts in the convection routines guarantees that the mass distribution with altitude is the same after the convection scheme as it was before it. DOWNDRAFT MASS FLUX FROM MOIST CONVECTION: These are the downdrafts which in the lower troposphere are about ? to 1/3 of the updrafts. They differ from the ÔcompensatoryÕ subsidence in that they can go through many levels in one time step, driven, for example, by evaporation of falling raindrops providing for cooling and increased density. SOLAR RADIATION HEATING RATE; THERMAL RADIATION COOLING RATE; TOTAL RADIATION COOLING RATE: The shortwave (solar radiation) and longwave (thermal radiation) cooling rates, in units of (10-2) degrees/day. Both of these are provided as positive numbers. The total radiation cooling rate, also a positive number, is therefore the thermal cooling minus the shortwave heating. It is expressed in energy units; so by doing the subtraction of the two previous diagnostics oneself, one can then have this difference in either degrees/day or W/(m2mb). TOTAL CLOUD COVER; SUPER SATURATION CLOUD COVER; MOIST CONVECTIVE CLOUD COVER: While one wouldnÕt necessarily think that the sum of the percentage super saturation clouds and convective clouds (each calculated in their own subroutine) would actually equal the total Ð if there were a supersaturation cloud and a moist convective cloud in the same box, it would seem there would be only one total cloud Ð in fact, the two do add to the total cloud cover. EFFECTIVE RELATIVE HUMIDITY: The air below a cloud that is evaporating moisture falling from the cloud will have a different specific humidity and a different temperature from the air surrounding the cloud. Since water holding capacity is an exponential function of temperature, the two specific humidity values should not be averaged first and then compared with the saturation value obtained from the (averaged) temperatures to obtain relative humidity. The relative humidity should be calculated first in each regime, and then averaged. This diagnostic does that; the relative humidity calculated this way is generally larger, especially at upper levels. It is termed ÔeffectiveÕ, because by taking into account the higher relative humidities beneath the cloud, it more accurately represents the ability of the air to evaporate precipitation. MOIST CONVECTIVE CLOUD FRACTION; LARGE-SCALE CLOUD FRACTION: (Newer diagnostics so their location may vary): As indicated for the sake of some physical processes, although for the sake of the radiation scheme, a cloud either fills all of the grid box or it is clear at each time step. WATER CLOUD COVER; ICE CLOUD COVER: This distinguishes the two particle types; the sum adds up to the total cloud cover, so it includes both supersaturation and moist convective clouds. WATER CLOUD OPTICAL DEPTH; ICE CLOUD OPTICAL DEPTH: Optical depth, or optical thickness, is a measure of transparency. Optical depth is defined as the negative natural logarithm of the fraction of radiation (e.g., light) that is not scattered or absorbed on a path, so the greater the optical depth, the less light gets through. Optical depth is dimensionless; the individual entries in the table are per 100 mb. EFFECTIVE WATER CLOUD PARTICLE SIZE; EFFECTIVE ICE CLOUD PARTICLE SIZE: The single scattering properties of clouds depend upon their cloud water or ice content, and the effective size of the particles. GCMs do not simulate particle size distributions, but an effective particle size can be parameterized from the water or ice content and temperature. TURBULENT KINETIC ENERGY: A parameterized calculation of the energy of turbulent eddies resulting from shear and buoyancy instabilities. HEATING BY LARGE SCALE CONDENSATION: There are two separate rainfall mechanisms in the model, large-scale condensation, of more importance in winter/baroclinic storms, and moist convective rainfall, more important in the tropics and warm conditions. This diagnostic shows the heat released when water vapor condenses in the large-scale condensation routine. HEATING BY TURBULENCE; CHANGE OF LATENT HEAT BY TURBULENCE: Turbulence moves both sensible heat and moisture around. These two diagnostics indicate those effects. CHANGE OF LATENT HEAT BY CTEI: Especially in the subtropics, cloud top entrainment instability is parameterized to occur due to mixing between saturated boundary layer air and unsaturated air from above the cloud top inversion. The instability, like turbulence, changes the moisture distribution and thus the latent heat. TOTAL HEATING BY MOIST CONVECTION (Q1-QR); TOTAL HEATING BY DEEP MOIST CONVECTION (Q1-QR); TOTAL HEATING BY SHALLOW MOIST CONVECTION (Q1-QR): Convection and convective clouds can change the (total derivative of) dry static energy (Q1) by altering the radiation (QR), by altering the sensible heat flux, or by inducing an excess of condensation over evaporation. These diagnostics calculate the heating by the last two of these processes. The convection scheme has both an entraining and a non-entraining plume, the former often associated with shallow convection, and the latter often leading to deep convection. TOTAL DRYING BY MOIST CONVECTION (Q2): Convection and convective clouds can also change the (total derivative of) specific humidity (Q2). This diagnostic indicates the energy equivalent of the moisture change (adding moisture produces more energy). MOIST CONVECTIVE EFFECTIVE CLOUD PARTICLE SIZE; LARGE-SCALE EFFECTIVE CLOUD PARTICLE SIZE: Equivalent to the water and ice effective particle size, this breaks it up into the different cloud generating subroutines. AVAILABLE POTENTIAL ENERGY: Associated with the temperature and temperature profile, available potential energy is the energy over and above what is needed to maintain a stable density profile. The validity of such a calculation at specific latitudes is somewhat questionable, since it requires defining a background stable profile, something that is usually done on at least a hemispheric scale. CHANGE OF U-WIND BY TURBULENCE; CHANGE OF U-WIND BY MOIST CONV: By mixing momentum, both turbulence and moist convection change the wind profile, here the U wind change is depicted. These then complement the other DU/DT diagnostics from the constant pressure section. U WIND AVERAGED OVER EAST PACIFIC; V WIND AVERAGED OVER EAST PACIFIC; VERTICAL VELOCITY FOR EAST PACIFIC; U WIND AVERAGED OVER WEST PACIFIC; V WIND AVERAGED OVER WEST PACIFIC; VERTICAL VELOCITY FOR WEST PACIFIC: The winds in the longitude sections corresponding to the east and west Pacific are given here, as a function of latitude and altitude. The rationale is that the meridional circulations spiraling away or towards the equator differ between the two regions, and also differ when ENSO events occur. NORTHWARD ELIASSEN-PALM FLUX; VERTICAL ELIASSEN-PALM FLUX; DIVERGENCE OF THE ELIASSEN-PALM FLUX: These diagnostics indicate how wave energy propagates, the values being normalized by the background wind. One can see that the waves in the troposphere generally propagate upward in the extratropics and then equatorward or poleward. The divergence diagnostic indicates where the wave energy is affecting the background winds; divergences speed up west winds, convergences reduce them, and where the values are close to zero, the waves are not affecting the mean flow at all. This diagnostic can be compared to the DU/DT by EP Flux Divergence Ð it is even in the same units Ð although they are on slightly different latitudinal grids, and this is not a constant pressure level diagnostic in the troposphere. The values are thus in closer agreement in the stratosphere, with its constant pressure levels. AMPLITUDE OF GEOPOTENTIAL HEIGHT FOR WAVE NUMBER 1,2,3,4: Wave number 1 indicates one trough and one ridge around a latitude circle Ð hence a wavelength of some 40,000 km (latitude dependent). Wave 2 is then half as large, etc. The amplitude in meters of these long planetary waves is evaluated at specific pressure levels by Fourier Analysis of the monthly average height fields. Thus they are both a combination of standing waves (not moving in space), and stationary waves (not changing in time) Ð as, often, are the observations against which they are compared. Since density decreases with altitude, if the energy is propagating freely, the amplitude will increase with altitude. When that stops happening, propagation is being impeded. PHASE OF GEOPOTENTIAL HEIGHT FOR WAVE NUMBER 1,2,3,4: The phase is defined as the longitude of the ridge of the wave, shown here in degrees W longitude (so 180¡ is the dateline, 165¡ is 15¡ east of the dateline, -165¡ is 15¡ west of the dateline). When waves are propagating vertically in a west wind regime (like mid-latitudes), the phase shifts westward with increasing altitude. If the phase is not changing with altitude, the wave is not propagating vertically. If the phase is shifting east with increasing altitude, the wave is propagating downward, as happens when a wave is being reflected. One can see that the shorter the wavelength, the less likely the wave is to propagate vertically (as was also indicated by the refraction indices). *****ZONAL AVERAGE LONGITUDE X ALTITUDE***** The following diagnostics all show various quantities around the longitudinal circle at specific latitudes. They are constant pressure diagnostics. ZONAL WIND (U COMPONENT) AROUND +/- 5; MERIDIONAL WIND (V COMPONENT) AROUND +/- 5 DEG; VERTICAL VELOCITY AROUND +/- 5 DEG; TEMPERATURE AROUND +/- 5 DEG; RELATIVE HUMIDITY AROUND +/- 5 DEG; MOIST CONVECTIVE HEATING AROUND +/- 5 DEG; TOTAL RADIATIVE COOLING AROUND +/- 5 DEG : The desire here was to be able to explore longitudinal variations in particular in the ENSO region of the East Pacific vs the West Pacific, as well as the Walker circulation, etc. Variations in all these parameters show up nicely in El Nino vs. La Nina situations, and also indicate what is going on in the Atlantic, etc. at those times. In combination with the latitudinal diagnostics for the East and West Pacific, they quickly depict how ENSO events are influencing many other regions. VERTICAL VELOCITY AT 50 N; TEMPERATURE AT 50 N; ZONAL WIND AT 50N: 50N was chosen because it is the region where the accuracy of tropospheric/stratospheric interaction is most easily observed, associated with the longitudinal variations of waves. The trough in the upper troposphere over the Pacific is ÔreplacedÕ by a ridge in the stratosphere (the ÔAleutian HighÕ), wave number 1 being the predominant stratospheric wave and predominant ridge feature in the stratosphere. This depiction will change during months of stratospheric warmings. VERTICAL VELOCITY AT 70 N; TEMPERATURE AT 70 N; ZONAL WIND AT 70 N: This latitude was chosen to look at troposphere/stratosphere interactions at high latitudes where the accuracy of the tropopause simulation becomes an issue, due to its influence on ozone intrusions from the stratosphere. *****WAVE POWER DIAGNOSTICS***** These diagnostics assess the power, wavelength and speed (period) of both tropical and extratropical waves. Wavenumber is given on the left, and period eastward or westward on the top. The diagnostics use the ÔMaximum Entropy MethodÕ which Òseeks to extract as much information from a measurement as is justified by the data's signal-to-noise ratioÓ. Here is it used to allow monthly data to provide the power of waves with periods of up to or more than a month, i.e., it is designed to make use of limited temporal information. It calculates an autoregressive time series to make predictions from prior data. The order of this series was originally calculated from the minimization of a parameter (which is shown below the table); subsequently, tests indicated that a value of two (i.e., two successive previous time periods) seemed to provide reasonable answers when evaluated with known waves. When the legend in the table reads 10 Day*(m/s)2, that implies the individual entries should be divided by 10. The value for the V wind at the equator at 300 mb and 50 mb should be 100D*(m/s)2, [instead it just listed (m/s)2], so those numbers need to be divided by 100. The cumulative VAR in the column second from the right is the sum of the power associated with the particular wavenumber. The speed of the waves can be determined by dividing the wavelength by the period of the wave. WAVE POWER FOR U,V NEAR 850 MB, 300 MB and 50MB AND EQUATOR: The diagnostic was set up here to look at tropical waves. Given the long wavelengths specified in the table (wavenumbers 1-9), the interest was in the planetary scale Kelvin Waves (wavenumbers 1-3, eastward moving periods 15-60 days), and Mixed Rossby-Gravity Waves (Wavenumbers 4-6, Westward moving Periods 4-6 days). Kelvin Waves have most of their energy in the U wind, while mixed Rossby-Gravity Waves have most of their energy in the V wind. Both waves can be seen clearly in the stratosphere (50 mb). The diagnostic could be changed, or added to, to provide results for higher wavenumbers, so as to look at easterly waves. WAVE POWER FOR PHI AT 922, 700, 500, 300, 100, and 10 MB AND 50 DEG N: Here the geopotential height of the wave is the diagnostic scrutinized. The diagnostic was set up to look at the speed of extratropical storms (wavenumber 6-9, eastward moving waves of 2 to 5 day period). One can see clear differences with model resolution, for example, with more energy in the faster moving waves at finer horizontal and vertical resolution. *****LATITUDE X LONGITUDE DIAGNOSTICS***** Most of these are readily interpretable, so comments here will be restricted to the few that are not. Each has a global average number in the lower left hand corner (next to the + signs). Scales are given at the bottom, and vary to some extent from one to another. Even if the scale doesnÕt say so, an X implies exactly 100% (as in the case of Ôland coverageÕ over all land), while a missing value is obviously 0% (as in the case of Ôland coverageÕ over the oceans). GROUND WETNESS (VEG ROOTS): This is the wetness in % relative to some chosen saturated value for the soil type (saturated values for soil types are actually a somewhat dubious concept). It is calculated down to the rooting depths, as determined from the vegetation type(s) specified for the grid box. Soil moisture below the rooting depth is largely unavailable for evaporation and has less influence on the atmosphere. GROUND TEMPERATURE: The ground has up to 6 levels of soil. This is the temperature of the first, shallow (2m) layer. SURFACE WIND SPEED, JET SPEED: These are the average velocities independent of direction. In other words, if the winds were 10 m/s from the east for half the month, and 10 m/s from the west for the other half, a true vector average would result in no wind velocity at all. Here the ÔspeedÕ would be listed as 10 m/s. The Ôjet stream levelÕ is ~200 mb. SURFACE WIND DIRECTION, JET DIRECTION: Here the vector strengths are taken into account to produce an average wind direction, clockwise from north. So a value of Ô9Õ, 90¡, indicates a wind from the west to the east, while ÔRÕ, 270¡, is a wind from east to west. TROP STATIC STABILITY: Uses the temperature at a defined tropopause altitude relative to the temperature of the modelÕs lowest layer to calculate the stability in ¡C/km. PEAK STATIC STABILITY IN PBL: Gives an indication of where the most unstable boundary layers are (in the tropics). NET RAD. OF PLANET: Short wave minus long wave energy at the top of the modelÕs atmosphere. BRIGHTNESS TEMP THRU WNDW: The temperature equivalent for the radiation at 10 microns at the top of the model atmosphere. CLR SKY INCIDENT SOLAR RADIATION, SRF: Value at the surface saved when cloud-free. NET THERMAL RADIATION, TOA: Net longwave radiation at the top of the atmosphere. NET HEATING AT GROUND: Takes into account the net radiation at the ground, and the sensible and latent heat fluxes removing energy from the ground, along with other minor factors. TOTAL NT DRY STAT ENRGY; NT DRY STAT ENR BY ST ED; NT DRY STAT ENR BY TR ED: The only energy transports that are defined on a latitude/longitude array, includes dry static energy (sensible heat + geopotential energy) by traveling and standing eddies, and by the total of eddies plus the mean circulation. TOTAL EARTH WATER: Includes water + ice down to the bottom of the model, including all soil layers. TAU>1 CLOUD COVER; CLOUD TAU=1 PRESSURE; CLOUD TAU=1 TEMPERATURE: The first diagnostic indicates what cloud cover occurs for clouds above a certain optical thickness (Tau of 1), while the other two indicate at what pressure and temperature these more ÔsubstantialÕ clouds first occur when looking down from above. At this optical thickness, they will influence the radiation to space and observations from remote sensing instruments. LOW LEVEL, MID LEVEL, HIGH LEVEL CLOUDINESS: Defined between different pressure levels; low level generally 1000-800 mb; mid level then up to 400 mb; high level above that. PBL HEIGHT: As calculated from the boundary layer scheme. GUSTI WIND: Parameterized to allow the model to, for example, raise dust to higher levels. CLOUD OPTICAL DEPTH (ISCCP); CLOUD TOP PRESSURE (ISCCP); LOW LEVEL CLOUDINESS (ISCCP); MIDDLE LEVEL CLOUDINESS (ISCCP); HIGH LEVEL CLOUDINESS (ISCCP); FRACTION OF TIME FOR ISCCP CLOUD: For comparison with ISCCP satellite data Ð which canÕt see below thick clouds, so the model utilizes the same routines as if it were also looking down from space. SW CLOUD RADIATIVE FORCING, TOA; LW CLOUD RADIATIVE FORCING, TOA: Breakdown to indicate how clouds are changing radiation at the top of the atmosphere in these different wavelengths. DEFORM. DRAG MOM FLUX; MTN WAVE MOM. FLUX; SHEAR WAVE MOM. FLUX; MC C=-10R, -20R, -40R MOM. FLUX: The model uses parameterized gravity wave momentum fluxes to affect the winds primarily above the tropopause. The fluxes are associated with model processes Ð the magnitude of wind deformation (e.g., in fronts), flow over mountains, wind shear between two levels and convective mass fluxes.These diagnostics give the source values of the parameterized momentum fluxes right above their generating region (primarily in the troposphere); they will vary with the specific processes. The parameterized shear waves are only generated when the shear exceeds a certain value similarly the flow over topography only generates waves for topography above a certain value, and the deformation must also exceed a certain limit. The -10R, -20R, etc. relate to the phase velocity relative to the background wind (equal magnitudes of +10R, +20R, +40R are also used); the magnitudes differ because only penetrating convection (through the 400 mb level) produces phase velocities greater than ?10R. PHASE SPEED OF SHEAR WAVE: The phase speed of the parameterized momentum fluxes associated with shear is the average speed of the two shearing levels. MC SOURCE WIND SPEED: The wind around which the convective momentum flux phase velocity varies (U,V +/-10,20,40 m/s). It is the background wind averaged over the convective depth. EXIT TOT. MOM. FLUX: The parameterized gravity wave momentum flux from all sources that reaches the top of the model Ð i.e., that has not been deposited by encountering background winds that either absorb it, or cause the waves to ÔbreakÕ and deposit some of it. In versions with a high model top (near the mesopause), all this momentum flux is allowed to break at the top. 850 mb, 700, 500, 300, 100, 30, 10, etc. MB HEIGHT: Monthly average values that can be calculated also from saved three dimensional fields of geopotential height. The two presentations donÕt necessarily have the same amount of ÔsmoothnessÕ to the fieldsÉ THICKNESS TEMP 1000-850, 850-700, 700-500, 500-300, 300-100, 100-30, 30-10, etc. The height difference between two pressure levels indicates the average temperature between them, determined here directly. [This diagnostic was used before we saved the three dimensional temperature fields, and still provides a quick assessment]. SEA LEVEL PRESSURE; SURFACE AIR TEMPERATURE: Originally, the maps had ÔoverstrikesÕ on the printer to highlight the land areas. That is no longer done; but they show up in this diagnostic as three rows of numbers for each land area point. *****CONSERVATION QUANTITIES***** The various processes that alter the modelÕs angular momentum, kinetic energy, potential energy, various ocean and lake quantities (mass, ice, salt) etc. For those quantities which should be conserved, the sum of the changes presented should be small relative to the instantaneous values, which are also shown. Note, like all of the diagnostics, these are not necessarily called at each time step, and so no exact conservation should be expected from the diagnostic routine. *****SPECTRAL ANALYSIS***** These diagnostics present an analysis of the energy of waves of different wavenumber is various regions: Southern Hemisphere, Northern Hemisphere, Equatorial and 45N; Troposphere, Lower, Middle, and Upper Stratosphere. The various columns represent the wavenumber (wavenumber zero is the zonal mean flow); Standing Kinetic Energy; Total Kinetic Energy (Standing + Traveling); Available Potential Energy; Changes by the Dynamic Subroutine (advection; coriolis force; pressure gradient force; total change by dynamics of kinetic energy; total change by dynamics of potential energy); changes in kinetic energy and potential energy by condensation/convection; changes in potential and kinetic energy by radiation; change in potential energy by surface friction; changes in kinetic and potential energy by the polar filter (and perhaps U,V filter); changes by daily restoration of mass in kinetic and potential energy associated with the surface sea level pressure filter; the kinetic energy on constant pressure surfaces (as opposed to the sigma level values given in column three); and the last values of kinetic and available potential energy at the end of the month. Looking at the difference of Ôlast valuesÕ from one month to the next indicates how much these quantities have changed during the month. There are several filters in use Ð a sea level pressure filter, to reduce B grid noise in the height field; a polar filter to allow for longer time steps near the pole; and, in Model E, a filter on the U and V wind fields. In addition to the values for the individual waves, the total of each of these quantities (over all the waves) is also shown. Therefore, given that zonal values (wavenumber zero) are also provided, Ôenergy boxesÕ for zonal and eddy, kinetic and available potential energy in the different model domains (e.g., troposphere, lower stratosphere, etc.) can easily be produced, compared with observations, and changes noted as climate changes. Zonal Available Potential Energy will change as latitudinal temperature gradients and vertical temperature profiles change; Eddy Available Potential Energy changes as longitudinal temperature gradients and vertical temperature profiles change. Eddy Kinetic Energy changes as waves are amplified or diminished with climate, and Zonal Kinetic Energy as the mean flow changes. The transformation from Zonal Available Potential Energy to Zonal Kinetic Energy (P-K in the Dynamics column for wavenumber 0) represents the generation of mean circulation cells (in the tropics, the Hadley Cell); the transformation from Eddy Available Potential Energy to Eddy Kinetic Energy (P-K in the Dynamics column at the bottom on the line for EDDY) indicates to a good extent the baroclinic process. K Adv (and K corr, in some numerical schemes) in the Dynamics column at the bottom on the EDDY line, indicates the flow from Zonal Kinetic Energy to Eddy Kinetic Energy or the reverse Ð the Ôbarotropic instabilityÕ process - and the total EDDY value should be the same with reverse sign of that given for wave number zero. The diagnostics also therefore allow for almost a complete direct assessment of Ôenergy flowsÕ in the different hemispheres and vertical regions. Only the transformation from Zonal Available Potential Energy to Eddy Available Potential Energy is not calculated, though that can be backed out by using the Ôlast valuesÕ from successive months or years to determine the changes, and then deduce what this last transformation must be. *****DIURNAL DIAGNOSTICS***** For four grid points (which have been chosen, but which can be varied), many values are collected each time step, and monthly average ÔdiurnalÕ diagnostics provided. They are printed in groups partly for visual convenience, and described that way below. (1) Incoming shortwave radiation, planetary albedo, ground albedo, shortwave radiation absorbed in the atmosphere, condensational heat release. (2) Surface pressure, potential temperature in layers 2-5. (3) Potential temperature in layer 1, Surface air temperature, temperature of the first layer of the ground, specific humidity in the 4th and 5th layers. (4) Specific humidity in layers 1-3, and then surface and Ôground levelÕ specific humidity (assuming saturation just above the ground at the temperature of the first layer of the ground). (5) Shortwave and longwave radiation absorbed at the ground, latent and sensible heat fluxes from the ground, net heating at the ground. (6) The geostrophic U,V and total wind at the top of the boundary layer (historically used in the boundary layer scheme), and the surface U and V winds. (7) Total wind velocity, cross isobar-angle flow angle (due to friction at the surface), the Richardson Number (a measure of local vertical temperature instability), the Rayleigh number (relates to whether energy transfer is primarily in the form of conduction or convection), and the surface drag coefficient for momentum. (8) Surface drag coefficients for heat and moisture, eddy diffusion coefficient in the boundary layer, depth of the boundary layer, and the frequency of dry convection. (9) Level to which dry convection (thermal instability) reaches, precipitation, evaporation, frequency of deep convection and shallow convection. (10) Cloud cover in layers 3-7. (11) Cloud cover in layers 1 and 2, total cloud cover, supersaturation cloud cover, convective cloud cover. (12) AOT Ð aerosol optical tau, e.g., the value associated with dust. Depending on the print-out, next may be shown the individual values for each of the ~30 days of the month for each of these quantities, so variances can be obtained. *****ENERGY HISTORY***** A print out of the zonal kinetic energy, eddy kinetic energy, standing eddy kinetic energy, zonal potential energy and eddy potential energy for each hemisphere for each day of the month. These are given for the NH and SH troposphere, low, mid and high stratosphere. One can thus see events such as stratospheric warmings occurring (sharp drop in zonal kinetic energy and increase in eddy kinetic energy in the middle stratosphere), as well as any other oscillations that may arise. *****RIVER FLOW DIAGNOSTICS***** Values are given for various rivers whose flow is calculated in the model. *****OCEAN DIAGNOSTICS***** (1) Ocean currents: Gulf Stream, Kuroshio Current and Antarctic Circumpolar Current values. (2) Overturning circulations: North Atlantic Deep Water, Pacific Deepwater, Antarctic Bottom Water (3) Northward transports of mass, potential enthalpy, and salt as a function of latitude in the Atlantic, Pacific, Indian and Global oceans. (4) Northward transport of heat in these different ocean basins and total ocean as a function of latitude by different mechanisms: (thermohaline) overturning circulation, Gent-McWilliams cross-isentropic parameterized flow, the (wind-driven) gyre circulation, and the total. (5) Northward transport of salt in these different ocean basins and total ocean as a function of latitude by these different mechanisms. (6) Diagnostics for the 14 ocean straits: vertical diffusion, transport of mass, transport of potential enthalpy, transport of salt. (7) Ocean mean quantities, as a function of level, of density, temperature and salinity. *****ISCCP-RELATED DIAGNOSTICS***** Cloud frequencies for different optical thickness at different pressure levels in different latitude zones (30-60N,S), (15-30N,S), (15N-15S) *****KEY DIAGNOSTICS***** For a quick look at the model results for each month, showing seasonal variations, trends over time in a run, etc.: global snow cover + ice cover, NH snow cover + ice cover, NH snow cover, NH ice cover, albedo, absorption by atmosphere, net radiation at P0 (top of atmosphere), net heat at surface (Z0), precipitation, sensible heat flux, latent heat flux, temperature of the ground, global temperature of the atmosphere, temperature gradients in the NH and SH, peak eddy kinetic energy in the NH and SH, zonal kinetic energy in the NH and SH, Eddy potential energy in the NH, zonal potential energy in the NH, Eddy Kinetic energy at the equator and mid-latitudes, Jet-stream value and latitude in the NH and SH, mean circulation streamfunctions in the NH, SH, and Max values, peak northward transfer of dry static energy by standing eddies, eddies and total; peak northward transport of static energy by standing eddies, eddies, total and latitude of the peak; peak northward transfer of zonal momentum by standing eddies, eddies, total and latitude of the peak. Also given is the Nino 3.4 index (positive values indicate El Nino, negative values La Nina). *****TRACER DIAGNOSTICS***** A general set of diagnostics has been developed for a set of 9 tracers that are used to diagnose model behavior, as well as for studies involving the particular tracers in their own right. The emphasis here will be on their use for model evaluation. The sources and sinks for these tracers can be varied; the particular set used for these nine involve estimated surface sources and simplified tropospheric and stratospheric chemistry. There are other diagnostics from more specialized tracers/species that can/should be added to this document, including aerosols and isotopes. The tracers used are SF6, Rn222, CO2, N2O, CFC11, 14CO2, CH4, O3, SF6 (alt). Their use for model development will be presented first, and then the various diagnostics. [Note Ð there is also a Kr85 tracer, not included here because it produced similar results to other tracers for interhemispheric transport]. There are actually two SF6 tracers Ð the first used observed growth with time, which was close to exponential. For the sake of determining the age of air, another SF6 tracer was added whose source distribution was the same, but instead was set to grow linearly with time (i.e., a constant source with time). Note: Do not confuse these tracers with the species that were included in the TCADI runs, in which more sophisticated chemistry/aerosols were embedded in the model. That list includes the following: Ox, NOx, ClOx, BrOx, N2O5, HNO3, H2O2, CH3OOH, HCHO, HO2NO2, CO, CH4, PAN, Isoprene, AlkylNit, Alkenes, Paraffin, Terpenes, isopp1g, isopp1a, isopp1g, isopp2a, apinp1g, apinp1a, apinp2g, apinp2a, HCl, HOCl, ClONO2, HBr, HOBr, BrONO2, N2O, CFC, codirect, stratOx, GLT, Water, DMS, MSA, SO2, SO4, H2O2_s, seasalt1, seasalt2, BCII, BCIA, BCB, OCII, OCIA, OCB, Clay, Silt1, Silt2, Silt3, Silt4, NH3, NH4, NO3p, SO4_d1, SO4_d2, SO4_d3, N_d1, N_d2, N_d3. Special diagnostics exist for them which are not covered here. Use of the Tracers for Model Development INTERHEMISPHERIC EXCHANGE TIME [ CFC11, SF6, CH4 (and Kr85)]: This is calculated with tracers that have an appreciable source only in one hemisphere (Northern Hemisphere). CFC-11 is a good example, and doesnÕt have additional complications, like tropospheric chemical loss. To estimate interhemispheric transport with the source in the N.H., calculate the (NH Ð SH) average concentrations, and divide by the flux across the equator. The simplest way to do this is to take the difference between the instantaneous values of the tracer in the two hemispheres at the end of the month (from the Tracer conservation diagnostics) and divide it by the Ôchange of tracer by DynamicsÕ (also in the conservation diagnostics - the magnitude is the same in each hemisphere, just opposite in sign). One can also use the values from the individual concentration and dynamic transport tables. SF6 can be used for this calculation as well, although since it is not destroyed in the stratosphere, some of the interhemispheric transport can occur at those levels, and then come back down into the S.H. troposphere. Methane can also be used, although it is affected by tropospheric chemistry, including different OH values in the Northern and Southern Hemispheres. Despite these differences, the different species give generally similar values, on the order of ~1 year. Note that a correction has to be made because there are a small sources in the S.H. for these species; Southern Hemisphere sources in our runs contribute 8% to the SF6 and CFC-11, and 1% for methane (considering both the sources and chemistry sinks), so one must increase the transport times accordingly. INTRAHERMISPHERIC TRANSPORT [CFC11, SF6, CH4, CO2 (and Kr85)]: These species can obviously also be used to gauge the relationship between the extratropical (source) concentrations, and those at the highest and lowest latitudes. This is especially useful when done seasonally, allowing the different dynamical mechanisms to be assessed. CO2 is of use as well to assess intrahemispheric transport in the Southern Hemisphere. TRANSPORT BETWEEN THE BOUNDARY LAYER AND THE FREE TROPOSPHERE [Rn222, CFC11, CO2]: For fast vertical transport, i.e., associated with convection, Rn222 is useful. With a source at the surface, and an exponential life time of 5.5 days (half life of 3.5 days), the vertical distribution of this tracer can be used to evaluate the mixing of convective schemes, and compared with some observations. For longer time-scale vertical transports one can use ratio of the upper troposphere to near surface values of any of the species with a surface source and no tropospheric chemistry or large trend (CFC11, CO2). BOUNDARY LAYER CONCENTRATIONS AND PROPERTIES [Rn222]: Another use of 222Rn, or in fact any of the tracers, would be to investigate the impact of finer vertical resolution on boundary layer mixing, both diurnal and seasonal. This could be done in conjunction with the four individual grid points for which many diurnal variation diagnostics are routinely kept. UPWARD TRANSPORT FROM THE TROPOSPHERE TO THE STRATOSPHERE [SF6, CO2, N2O, CFC-11 CH4]: SF6 is particularly useful because it is a passive tracer with no chemical loss in the troposphere or stratosphere. A straightforward approach is to calculate the ratio of stratospheric/tropospheric concentration. The model output includes a definition of the tropopause at each latitude, and it is used in the budget pages but the level is not printed out. Alternatively, a simple yet sufficient approach, for annual average output, is to use 100 mb in the tropics, 200 mb at mid-latitudes, and 300 mb at high latitudes; in the extratropics, the dividing line would be at levels of potential vorticity close to 2 x 10-6 m2 s-1 K kg-1 (from the potential vorticity diagnostic). One can then compare this stratosphere - troposphere difference to the total vertical transport through the tropopause, to get a troposphere/stratosphere exchange time analagous to the interhemispheric exchange time. One can also use other species with tropospheric sources and compare their stratospheric/tropospheric differences or ratios, i.e., CO2, N2O, CFC-11, CH4. Although they each have complications (atmospheric chemistry or seasonal sources) they generally tell a similar picture concerning the change of transport upward into the stratosphere, be it with respect to climate change or model configuration. TROPSOSPHERIC/STRATOSPHERIC AGE OF AIR [(SF6)]: The age of air is calculated as a lag between the concentration at any level and the global first-layer concentration (for troposphere/stratosphere age of air), or the level below the tropopause (for stratospheric age of air). This is best done with the SF6 tracer whose growth is linear with time (SF6_c in the model output). TRANSPORT WITHIN THE STRATOSPHERE [SF6, CFC-11, N2O, CH4]: Since the source of these species in the stratosphere is upwelling through the tropical tropopause, their distribution away from that region in the lower stratosphere is indicative of the effect of tracer transport Ð as long as one is below the level in which chemistry dominates transport. Peak chemical loss occurs at ~43 mb for CFC-11, ~20 mb for N2O, and ~10 mb for CH4. For CO2 and SF6 with no chemical loss, even transport to the mesosphere can be assessed Ð again, both between different model formulations, and due to climate change. In particular, the latitude spread of age of air contours is indicative of the ÔleakyÕ stratospheric pipe. One can quantify this aspect by relating the ratio of SF6 in the tropics to the value at 30 N/S for various pressure levels in the low to middle stratosphere. TRANSPORT FROM THE STRATOSPHERE TO THE TROPOSPHERE [14C, SF6, CO2, O3]: Bomb-produced 14C can be used to calculate the stratospheric residence time, and hence transport down into the troposphere. The source has a release date of October 1963. Some observations are available of its subsequent concentration in the stratosphere as a function of time. The residence time in the stratosphere is calculated by fitting a least mean square line to the natural logarithm of the concentration as a function of time, and can be done for the short-term (a period of a few months), and longer term (a period of two to three years). One can also directly compare the downward transport through the tropopause for this species, as well as for other species without stratospheric sinks (SF6, CO2). And direct fluxes of ozone down into the troposphere can be compared with observations/models, although due to its stratospheric source that is also greatly influenced by transport within the stratosphere, as well as stratospheric photochemistry. Observed ozone fluxes into the troposphere are generally in the range of 400Ð600 Tg yr-1. [A non-tracer assessment of this transport is to use the total vertical transport of potential vorticity diagnostic, the values through the tropopause at latitudes outside of the tropics]. DETERMINATION OF STRATOSPHERIC OR TROPOSPHERIC ORIGIN OF AIR MASSES [(N2O, CH4, O3, CFC11, CO2, SF6, O3)]: Simultaneous observation of multiple tracers is used for this purpose. Species with stratospheric photochemical sinks have reduced ratios relative to conservative species when the air comes from the stratosphere. So one could compare, for example, N2O with CO2 or SF6 as a function of altitude, to determine the predominance of air from one of these regimes or the other. Similarly, high O3 air in the upper troposphere typically has a stratospheric origin, if sufficiently away from geographic regions of pollution. [For another differentiation, air with high potential vorticity comes from the stratosphere]. For more information on the use of these on-line tracers for model development and climate change impact, discussions of the sources and sinks, etc. see the following references: Rind, D., and J. Lerner, 1996: Use of on-line tracers as a diagnostic tool in general circulation model development: 1. Horizontal and vertical transport in the troposphere. J. Geophys. Res., 101, 12667-12683, doi:10.1029/96JD00551. Rind, D., J. Lerner, K. Shah, and R. Suozzo, 1999: Use of on-line tracers as a diagnostic tool in general circulation model development: 2. Transport between the troposphere and stratosphere. J. Geophys. Res., 104, 9151-9167, doi:10.1029/1999JD900006. Rind, D., J. Lerner, and C. McLinden, 2001: Changes of tracer distributions in the doubled CO2 climate. J. Geophys. Res., 106, 28061-28079, doi:10.1029/2001JD000439. Rind, D., J. Lerner, J. Perlwitz, C. McLinden, and M. Prather, 2002: Sensitivity of tracer transports and stratospheric ozone to sea surface temperature patterns in the doubled CO2 climate. J. Geophys. Res., 107, no. D24, 4800, doi:10.1029/2002JD002483. Rind, D., J. Lerner, J. Jonas, and C. McLinden, 2007: The effects of resolution and model physics on tracer transports in the NASA Goddard Institute for Space Studies general circulation models. J. Geophys. Res., 112, D09315, doi:10.1029/2006JD007476. Rind, D., 2010: The use of tracers as diagnostics for model development. In ECMWF Seminar on Diagnosis of Forecasting and Data Assimilation Systems, 7-10 Sep. 2009. European Centre for Medium-Range Weather Forecasts, 201-220. *****LATITUDE X ALTITUDE TRACERS***** (for each of the 9 specieS) TRACER CONCENTRATION: presented as a function of latitude and altitude (sigma coordinates) for each of the 9 species. TOTAL NORTHWARD TRANSPORT OF TRACER MASS; NORTHWARD TRANS. OF TRACER MASS BY EDDIES: the total includes both the eddy and mean circulation contributions. TOTAL VERTICAL TRANSPORT OF TRACER MASS; VERTICAL TRANS. OF TRACER MASS BY EDDIES: again, the total includes the eddy plus mean circulation components. CHANGE OF TRACER MASS BY MOIST CONVECTION: includes both the upward movement due to the rising plume, and transport by downwelling and compensatory subsidence. CHANGE OF TRACER MASS BY LARGE-SCALE CONDENSE: for species capable of being rained out. CHANGE OF TRACER MASS BY TURBULENCE/DRY CONVECTION: predominantly in the boundary layer. *****SOURCES/SINKS FOR TRACERS***** SF6 CFC-GRID SOURCE, LAYER 1: specified latitudinal-average input values as a function of time and latitude for these species; in this representation they use the same (latitude x longitude) source function, represented by industrialization. LOSS OF RADON-222 BY DECAY: latitude x altitude, time-dependent calculation. RADON-222 SOURCE, LAYER 1: latitudinal average source function associated with soils. CO2 FOSSIL FUEL SOURCE; CO2 FERTILIZATION SINK; CO2 NORTHERN FOREST REGROWTH SINK; CO2 FROM LAND USE MODIFICATION; CO2 ECOSYSTEM EXCHANGE; CO2 OCEAN EXCHANGE: The various latitudinal average CO2 sources in the first atmospheric layer. CHANGE OF N20 BY RESETTING TO 462.2d-9: when out-of-balance, change put into first atmospheric layer. CHANGE OF N2O, CFC-11 BY CHEMISTRY IN STRATOS: latitude x altitude representations. CHANGE OF CFC-11 BY SOURCE: Latitudinal-average of input into first atmospheric layer. CHANGE OF 14CO2 by SINK: 14CO2 input in the stratosphere, and removed from the atmosphere when it reaches the first atmospheric layer. CH4 SINK DUE TO SOIL ABSORPTION; CH4 TERMITE SOURCE; CH4 OCEAN SOURCE; CH4 FRESH WATER LAKE SOURCE; CH4 MISC_GROUND SOURCE; CH4 WETLANDS+TUNDRA SOURCE; CH4 ANIMAL SOURCE; CH4 COAL MINE SOURCE; CH4 GAS LEAK SOURCE; CH4 GAS VENTING SOURCE; CH4 MUNICIPAL SOLID WASTE SOURCE; CH4 COAL COMBUSTION SOURCE; CH4 BIOMASS BURNING SOURCE; CH4 RICE CULTIVATION SOURCE: all the methane contributions, diagnostics show latitudinal average values put into first atmospheric layer (as in the other sources, there are longitudinal variations). CHANGE OF CH4 BY CHEMISTRY IN TROPOSPHERE; CHANGE OF CH4 BY CHEMISTRY IN STRATOS; TOTAL CHANGE OF CH4 BY CHEMISTRY : Latitude x altitude representations. CHANGE OF O3 BY DEPOSITION IN LAYER 1: In this representation, ozone is removed at a rate by deposition from the lowest layer. CHANGE OF O3 BY CHEMISTRY IN STRATOS; CHANGE OF O3 BY CHEM PROD. IN TROPOSPHERE; CHANGE OF O3 BY CHEM LOSS IN TROPOSPHERE: latitude x altitude representation, breaking the tropospheric atmospheric chemistry change into production and loss values. TOTAL CHANGE OF O3 BY CHEMISTRY AND DEPOSITION: Combining the previous tables. SF6_c CFC-GRID SOURCE, LAYER 1: Alternate source variation with time (constant source with time, resulting in linear growth with time). *****LATITUDE X LONGITUDE FOR TRACERS***** TRACER TOTAL MASS, TRACER AVERAGE, TRACER AT SURFACE, TRACER BY VOLUME AT SURFACE: for each of the nine tracers. SOURCE DISTRIBUTIONS FOR THE VARIOUS TRACERS: those put from the surface into Layer 1. ATMOSPHERIC SOURCES AND SINKS, INCLUDING CHEMISTRY FOR THE VARIOUS TRACERS: from simplified chemistry schemes, for both the troposphere and stratosphere, or the proscribed losses. ATMOSPHERIC CONCENTRATIONS OF THE VARIOUS TRACERS, FOR EACH MODEL LEVEL: for all 9 tracers. *****CONSERVATION QUANTITIES FOR TRACERS***** INSTANTANEOUS VALUE OF THE TRACER: At the last time step of each month as a function of latitude. CHANGE OF TRACER BY DYNAMICS, CONDENSATION, LAND SURFACE, SURFACE, FILTER, OCEAN, VARIOUS INDIVIDUAL SOURCES, TROPOSPHERIC CHEMISTRY, STRATOSPHERIC CHEMISTRY, SUM OF CHANGES: As a function of latitude. The sum of the changes should equal the difference between the Ôinstantaneous valueÕ from the previous month, and the value for this month. 5