1

DISTRIBUTION STATEMENT A. Approved for public release; distribution is unlimited.

Capturing the Stratosphere’s Influence on Seasonal and Intraseasonal Predictability in a State-of-the-Art

Navy Global Environmental Model (NAVGEM)

Stephen D. Eckermann Geospace Science & Technology Branch

Code 7631, Space Science Division Naval Research Laboratory

Washington, DC 20375 Phone: (202) 404-1299 Fax: (202) 404-7997 email: stephen.eckermann@nrl.navy.mil

Award Number: N0001412WX21321

http://www.nrl.navy.mil/ssd/ LONG-TERM GOALS Recent research has revealed that the stratosphere influences medium- and long-range weather prediction, sometimes strongly (NAS 2010). The North Atlantic Oscillation (NAO) – one of the most prominent modes of intraseasonal tropospheric variability extending from the subtropical Atlantic to the Arctic (Hurrell et al. 2003) – has been recognized only within the past decade as one regional manifestation of a larger hemispheric phenomenon, known synonymously as the Arctic Oscillation (AO) or Northern Annular Mode (NAM). The NAM extends continuously into the stratosphere and mesosphere, and an analogous deep Southern Annular Mode (SAM) occurs in the southern hemisphere. NAM/SAM anomalies often appear first in the upper stratosphere or mesosphere, then descend gradually over a period of weeks, sometimes reaching the surface where they change weather patterns throughout the polar region (Baldwin and Dunkerton 2001; Coy et al. 2011). Descending stratospheric NAM/SAM anomalies also play a pivotal role in controlling the response of high-latitude weather to the El-Niño/Southern Oscillation (ENSO) in the tropics (Bell et al. 2009; Ineson and Scaife 2009), while the tropical stratosphere and mesosphere may also impact tropical seasonal prediction through an improved Madden-Julian Oscillation (Weare et al. 2012). These examples, and others like them, point to the important role that the overlying stratosphere-mesosphere system can play in presaging and regulating large-scale global surface weather changes over periods of weeks to months (e.g., Baldwin et al. 2003), prompting recent reports from the World Climate Research Programme (WCRP 2008) and the National Academy of Sciences (NAS 2010) that note (quote) “the stratosphere’s potential to improve seasonal forecasts is largely untapped.” Thus the long-term goals of this project are to tap the potential of an improved stratosphere and mesosphere for seasonal prediction by developing and testing new stratospheric and mesospheric modeling, prediction and data assimilation capabilities, all specifically designed to improve long-range prediction capabilities of the Navy Global Environmental Model (NAVGEM).

2

Table 1. Summary of Year 1-3 progress on Tasks R1 and R2, followed by a list of peer-reviewed publications arising from this research and the specific project task areas that they address.

R1 Tasks R2 Tasks

R1a Completed R2a Partial Progress

R1b Partial Progress R2b Completed

R1c Significant Progress R2c Significant Progress

R1d Completed R2d Significant Progress

R1e Completed R2e Partial Progress

R2f Partial Progress

Peer-Reviewed Publication Tasks Addressed Hoppel et al. (2013) R1e

Eckermann et al. (2014) R1a Coy and Reynolds (2014) R2d R2e

Tripathi et al. (2015) R2b

McCormack et al. (2015) R1a R1d R2a R2c R2d R2e R2f

Eckermann et al. (2015) R1d Tripathi et al. (2015) R2b

OBJECTIVES The overarching objective of this ONR-sponsored research is to gain an improved understanding of how the stratosphere and mesosphere in a state-of-the-art numerical weather prediction (NWP) system affect atmospheric prediction on time scales from days to months. To achieve this, our research focuses on the following scientific questions: 1. What are the fundamental dynamics and dominant physical coupling pathways governing the

stratosphere-troposphere interaction that are most relevant for atmospheric prediction on time scales from days to months?

2. Which physical and dynamical processes in the forecast model are important in controlling this deep vertical coupling, and how sensitive is forecast skill to details in their numerical implementation?

APPROACH Our primary tool is NAVGEM, the Navy’s global NWP system, comprising a semi-implicit semi-Lagrangian (SISL) forecast model coupled to the four-dimensional variational (4DVAR) NRL Atmospheric Variational Data Assimilation System – Accelerated Representer (NAVDAS-AR). Our approach to utilizing NAVGEM for this research is guided by the following recommendations of WCRP (2008) and NAS (2010) to address knowledge gaps in our current understanding of coupled troposphere-stratosphere predictability:

3

R1. Models must extend to at least 0.01 hPa (~80 km altitude) with a full range of appropriate physical parameterizations, so as to properly predict stratospheric and mesospheric variability;

R2. Research should then focus on the extended system’s ability to reproduce and delineate the poorly understood connections between stratospheric and tropospheric circulations.

Figure 1. Schematic depiction of the altitude range and order of development of various model-dependent processes affecting the heat budget (and hence temperature biases) in the NAVGEM

stratosphere and mesosphere (see text for further details).

R1 Tasks To address R1, our project is progressively augmenting the NAVGEM SISL forecast model and DAS in ways that improve its ability to predict the integrated troposphere-stratosphere-mesosphere system, as outlined in tasks (a)-(e) below. These tasks are described, justified and linked scientifically in detail in the original science proposal.

(a) SISL model tests using additional vertical layers and resolution

(b) Improved radiative heating and cooling rates for the SISL model’s stratosphere and mesosphere

(c) Improved ozone and water vapor photochemisty for the upper stratosphere and mesosphere

(d) Parameterizations of subgrid-scale gravity-wave drag for the stratosphere and mesosphere

(e) Assimilation of stratospheric and mesospheric satellite observations

R2 Tasks To address R2, we perform the following component tasks using the augmented NAVGEM model, as progressively developed via successful execution of the R1 tasks listed above. Again, these tasks are described and scientifically motivated in detail in the original science proposal.

4

(a) Bias identification and parameterization correction/improvement

(b) Forecast-assimilation experiments supporting international research projects on predictability

(c) Realistic tropical QBO (quasi-biennial oscillation) and SAO (semiannual oscillation)

(d) Prediction of stratospheric sudden warmings (SSWs) and stratospheric NAM/SAM anomalies

(e) Tropospheric influences of stratospheric NAM/SAM anomalies

(f) Tropospheric coupling to the stratospheric QBO

WORK COMPLETED The progress on project tasks in the first 3 years of research under this DRI are summarized in Table 1. Significant progress was made in nearly all of the R1 tasks, with tasks R1a, R1d and R1e completed, as documented in peer-reviewed scientific publications for Task R1e (Hoppel et al. 2013), Task R1a (Eckermann et al. 2014), and Task R1d (Eckermann et al. 2015). The collaborative forecasting tasks under the aegis of the international SNAP (Stratospheric Network for Atmospheric Prediction) consortium (Task R2b) was also completed through to publication (Tripathi et al. 2015a 2015b). Work was also completed on a first demonstration of self-generated QBO circulations in a Navy global model and the resultant tropical-extratropical teleconnections to the high-latitude winter that in turn modulate late-winter Arctic sea-level pressure and associated weather through descending NAM anomalies (McCormack et al. 2015), yielding progress in Tasks R2c, R2d, R2e, and R2f. Furthermore, key technical aspects of the completed R1 tasks, reflected as new NAVGEM model capabilities such as improved water vapor chemistry and quality control (Task R1c), new hybrid-coordinate vertical layers (Task R1a), and a stochastic parameterization of nonorographic gravity-wave drag (Task R1d), have all transitioned to operations during 2015 as part of NAVGEM 1.3, as documented in the Validation Test Report (VTR) of Whitcomb et al. (2015). RESULTS Here we summarize our Year 4 research that focused on uncompleted research tasks and their associated objectives as summarized in Table 1. Heat Budget of the NAVGEM Stratosphere and Mesosphere The remaining R1 task for which significant further research was required was Task R1b to improve parameterizations of radiative heating and cooling rates for the SISL model’s stratosphere and mesosphere. Likewise Task R2a, to identify and correct biases in the NAVGEM stratosphere and mesosphere, also required further work. This year both of these issues were addressed via a detailed scientific study of the heat budget of the NAVGEM stratosphere and mesosphere.

5

Figure 2. Zonal-mean temperature differences (contours in K: see color bar) as function of altitude

(pressure in hPa) and latitude (deg.) between two NAVGEM forecast runs that are otherwise identical apart from the level of decentering ε (values of 0.1 and 0.6). Note large zonal-mean

temperature differences >10-20 K between the two runs at upper levels. Figure 1 presents a simplified schematic summary of the complex interacting processes that we have identified as impacting systematic cold biases in the upper layers of the NAVGEM model. Initial work in prior years focused primarily on the physical parameterizations, depicted in orange, red and blue. These experiments revealed that cold biases varied in unexplained ways in response to changes in model resolution, time step and small-scale numerical dissipation, an effect that made the model cold bias a “moving target.” This effect was studied and eventually attributed to some previously undiagnosed issues in the NAVGEM SISL dynamical core. Two previously unappreciated effects were identified. First, the new prognostic equation for potential temperature in the SISL model removes a previous mode of instability within the discretized primitive equations by solving the thermodynamic equation in terms of a perturbation potential temperature variable ϴ’, defined in terms of deviations from an imposed one-dimensional reference state profile Θ�, as explained in section 2.3 of the NAVGEM 1.3 Validation Test Report (Whitcomb et al. 2015).

6

Second, and relatedly, a so-called decentering algorithm was added to NAVGEM, which applies a targeted form of numerical dissipation that suppresses an unstable growing resonance mode over orography. Rivest et al. (1994) showed that these spurious oscillations can be very effectively suppressed by the addition of a small amount of damping, quantified by a so-called decentering or uncentering parameter, ε, where 0 ≤ 𝜖 ≤ 1. This decentering acts on the averaging of state-dependent terms in the SISL discretization of the primitive equations. For a state parameter F, the mean value along an SL trajectory is now computed as

𝐹 = �1+𝜖2�𝐹𝑎� (𝑡 + ∆𝑡) + �1−𝜖

2�𝐹𝑑(𝑡 − ∆𝑡) . (1)

Here t is time, Δt is the model time step, subscripts “a” and “d” denote the SL arrival and departure points, respectively, and the tilde represents terms prior to application of any time filter to suppress growing computational modes (Cordero and Staniforth 2004).

The decentering defined in (1) is designed to suppress fast unphysical gravity-wave modes near the model’s truncation scale (Rivest et al. 1994), and so should have little-to-no effect on zonal-mean state parameters such as temperature. While our NAVGEM experiments show that this holds well throughout the troposphere, in the stratosphere and mesosphere we found a significant dependence of the zonal-mean temperature on the value of ε, as shown in Figure 2. Furthermore, routine NAVGEM modeling experiments were using large decentering values (ε=0.6) since the warming trend evident in Figure 2 from increasing ε reduced mean temperature biases. Yet kinetic energy spectra (not shown) revealed that such large values of ε significantly dissipated energy at large wavenumbers, which is an undesirable effect of large ε.

Figure 3. Schematic depiction of new parameterization of shortwave solar heating rates (red), its

relation to the existing RRTMG parameterization (black), and new parameterizations of upper-level heating due to exothermic photochemical reactions (yellow) and radiative airglow losses (green).

See text for further details.

7

These problems have been resolved by first recognizing that the prognostic ϴ’ variable must be small at all altitudes for these new solution methods to work as designed. On investigating, we found that the current Θ� reference state in NAVGEM was excessively cold at upper altitudes, leading to large |ϴ’| values at upper levels that in turn led to model errors, manifested as excessive cold bias and an unphysical dependence of zonal-mean quantities on the ε value. After replacing the current reference profile in NAVGEM with a new profile that better matches zonal-mean prognostic ϴ values at all model altitudes, we found significantly reduced sensitivity of zonal-mean temperature to ε, which in turn allowed us to choose a more desirable (small) value of ε=0.1 to retain realistic kinetic energy at large wavenumbers while still suppressing orographic resonance instabilities. Likewise, sensitivities of temperature bias to time step were largely eliminated with this change, meaning that the dynamical core sources of cold bias, depicted in green in Figure 1, have now been largely eliminated, such that cold bias at upper levels is now no longer a “moving target” and can be addressed in terms of missing physics. Shortwave Radiative Heating Rates With the elimination of dynamical core cooling, we next addressed radiative heating rates at upper levels. Radiative heating rates in NAVGEM are currently parameterized using the Rapid Radiative Transfer Model for GCMs (RRTMG: Clough et al., 2005). The RRTMG parameterization is based on assumptions of local thermodynamic equilibrium (LTE) in which all absorbed solar energy is assumed to convert, essentially instantaneously, into local atmospheric heating. While this is a very good approximation at low altitudes where high atmospheric pressures lead to rapid collisional interactions that achieve an LTE-like state, at higher altitudes the low atmospheric pressures cause LTE to break down. This in turn requires major changes to the way in which absorbed solar energy is parameterized in terms of net atmospheric heating effects. After discussions with personnel from Atmospheric Environmental Research (AER), the company responsible for issuing and moderating the RRTMG code for the wider modeling community, it became clear that AER had no funding avenues or proposal plans to extend the RRTMG to include the additional absorption bands and non-LTE effects needed for parameterizing heating rates at upper altitudes. Hence we were forced to consider other parameterization strategies for NAVGEM. Our parameterization approach to the problem is summarized in Figure 3, with the new solar heating-rate parameterization depicted in red. Since LTE breaks down, this new parameterization must delineate additional flows of absorbed shortwave solar photon energy into two additional forms: an airglow loss pathway, depicted in green in Figure 3, and a chemical pathway, shown in yellow in Figure 3. The chemical pathway, leading to heating via exothermic chemical reactions, is discussed later. Our upper-level radiative heating-rate parameterization, depicted in red in Figure 3, is a new standalone module that focuses on the absorption of ultraviolet (UV) and visible solar radiation in the 200-800 nm range, which is most relevant for heating from the stratosphere up though the mesosphere and lower thermosphere (MLT).

8

Table 2. Physical constants of the new upper-level shortwave radiative heating-rate parameterization. Unless otherwise noted values are either original values from Strobel (1978) or original/updated values from Zhu (1994). εD and quantum yield values are derived in this work

and/or from indicated sources. sb values are adapted from fits in Zhu et al. (2003).

The various absorption bands and their wavelength ranges are summarized in columns 1 and 2 of Table 2. The parameterization includes bands for the major O3 absorption wavelengths, similar to what is in the RRTMG, although here the band resolution is finer and focuses on peaks in the absorption cross sections. A new feature absent from the RRTMG is O2 absorption bands at wavelengths shorter than 240 nm, spanning the Herzberg continuum, the Schumann-Runge bands (SRB), the Schumann-Runge continuum (SRC) and the solar H Lyman-α emission. Prescribed solar flux values and absorption cross sections within all these bands are listed in columns 4 and 5, respectively, of Table 2. Our heating rate solutions assume constant values of cross section within the Chappuis band, Hartley band, Herzberg continuum and SRB. By contrast in the Huggins bands and SRC we assume an exponential variation of cross section with wavelength, which better fits actual cross sections within these wavelength intervals. The resulting heating rate expressions are similar to those first derived by Lindzen and Will (1973) and Strobel (1978), then later updated and extended by Zhu (1994). Our Lyman-α heating-rate parameterization is based on the analytical expressions of Chabrillat and Kockarts (1997, 1998) for absorption within this band, which take account of wavelength and temperature variations in absorption cross section across the band. Another addition recognizes that at these shorter UV wavelengths we can no longer assume a constant solar insolation, but instead must now account for variations in solar flux by band in response to

9

changes in solar activity. Following Zhu et al. (2003), we have linearly regressed observed solar fluxes within each band to the standard metric of solar activity based on the 10.7 cm solar radio flux index (f10.7) as expressed in solar flux units (SFU: 10−22 W m−2 Hz−1). The resulting linear fits are expressed in terms of a normalized linear slope index sb shown in the far-right column of Table 2 for each band, which allows the standard fluxes in column 4 of Table 2, now assumed to hold for a typical f10.7 = 145 SFU, to be scaled in response to the input f10.7 of the day. The increase in sb with decreasing band wavelength in Table 2 reveals the progressively greater role of solar activity in modulating solar fluxes at shorter UV wavelengths. Another important new feature of the scheme is quantification within each band of the components of absorbed photon energy going into atmospheric heating Q, photochemical products D and airglow losses A. If E=Q+D+A is the total photon energy within the band, then we quantify this partitioning in terms of a so-called chemical efficiency εD = (E-D)/E and an airglow efficiency εA=(E-A)/E. We have parameterized the airglow efficiency εA based on model simulations and fits presented by Mlynczak and Solomon (1993) for the Huggins and Hartley O3 bands and for the SRC O2 region. Our tests (not shown) indicated that these parameterized losses are generally no more than ~10%. A far more important temporary sink for absorbed solar energy is in the products of photochemical reactions produced by UV photoloysis, due to the energy required to break the strong bonds in O2 and O3 molecules and to excite atomic oxygen from the ground-state O(3P) to the excited singlet-delta state O(1D). Quantifying εD in each band requires (a) knowledge of bond dissociation energies for the major photolysis reactions within each band, and (b) the so-called quantum yield of monatomic oxygen in its excited O(1D) relative to the ground-state O(3P). We have performed these calculations for all our bands based on observations of dominant reaction pathways as a function of wavelength and laboratory measurements of quantum yields for photodissociation of both O3 and O2 as a function of photon wavelength. The results of these εD calculations are given in Table 2 and reveal generally low values, indicating that a large fraction of absorbed solar energy is being locked in chemical products and is only liberated as heat via subsequent exothermic recombination reactions, the latter requiring a separate parameterization (see yellow box in Figure 3). This chemical heating contribution must be removed here from the direct solar heating to avoid “double counting” leading to an unbalanced (overheated) thermal energy budget at upper levels in NAVGEM.

10

Figure 4. Red profiles show daytime values of (a) temperature (K) and (b) ozone mixing ratio (9.6

μm), derived from version 2.0 retrievals of SABER observations from March-April 2003-2012 at the equator (5◦S–5◦N), which are used in tests of shortwave heating code. Other curves in (a) show

corresponding nighttime (blue) and April diurnal (black) means, revealing large day-night temperature asymmetries above ~70 km altitude due to the diurnal tide. Blue curve in (b) shows

corresponding daytime ozone from the SABER 1.27 μm retrieval. Early results using this upper-level shortwave parameterization are encouraging. We have tested its performance using prescribed vertical profiles of daytime temperatures and ozone mixing ratios derived from observations from the SABER (Sounding of the Atmosphere Using Broadband Emission Radiometry) instrument on NASA’s TIMED satellite (Remsberg et al. 2008). We derived daytime means for March-April over the years 2003-2012 inclusive (red curves in Figure 4). The blue curve in Figure 4a shows the corresponding nighttime temperature profile, while the black curve shows the diurnal mean temperature for April. The large temperature variability above ~70 km altitude is due to the diurnal tide. For O2 we assume a constant mixing ratio with height of 0.20949 and derive O2 number and column densities from this mixing ratio and local atmospheric density. The resulting heating-rate profiles generated by our new parameterization are displayed as the thick solid black curves in Figure 5, with results on the top row for an overhead Sun (cos𝜒 = 1 where χ is the local solar zenith angle) and results on the bottom row for the Sun low in the sky (cos𝜒 = 0.2). Results in the left column show heating rates with efficiencies set to unity (ε=εD=εA=1), typical of a standard LTE-based parameterization such as the RRTMG. The dashed and dotted gray curves show the contributions to the heating rate from the O3 and O2 absorption bands, respectively. These profiles show that heating rates are dominated by O3 photolysis below ~65 km, but that above this level, heating due to O2 photolysis quickly becomes competitive with, and ultimately larger than, O3 absorption. This confirms both the adequacy of the O3-only formulation of RRTMG below ~60 km, and its inadequacy for specifying heating rates above 60 km.

11

Figure 5. Shortwave heating rates (black curve) versus altitude for mean temperature and daytime

ozone profiles in Figure 4, as well as contributions (see key, bottom right) due to O2 and O3 absorption (gray broken curves) and individual absorption bands in Table 2 (colored curves).

Results are shown for (left column) an LTE assumption of ε = εA = εD = 1 at (a) cos χ = 1 and (b) cos χ = 0.2, and (middle column) nonLTE assumptions (0 < ε = εA+εD−1 < 1) for (c) cos χ = 1 and (d)

cos χ = 0.2. Panel (e) shows the nonLTE-to-LTE ratio of component heating-rate profiles in panels (c) and (a) for cos χ = 1.

The colored curves show the heating rate contributions from all the individual bands listed in Table 2. The relative shapes, peaks and values for these band heating rates compare favorably to a similar presentation in Figure 1 of Strobel (1978) for equatorial equinox, and to the heating rate profiles for equatorial January shown by Nissen et al. (2007) and Sukhodolov et al. (2014). The middle column of Figure 5 shows the same presentation but with efficiencies now activated (i.e. 0< εD,εA <1). While the normalized form of these profiles look very similar to those in the left panels, note that the total heating rates are now substantially reduced. To quantify and illustrate these reductions better, Figure 5e plots the ratios of the cos χ = 1 heating rates with efficiencies activated (Figure 5c) relative to those with efficiencies deactivated (Figure 5a), quantifying the net effects of airglow and chemical efficiency terms on local shortwave heating rates. From ~40-90 km altitude the ratio of the total heating rates stays at a roughly constant value of 25%, indicating that only about a quarter of the total solar UV radiation absorbed through these altitudes appears immediately as local atmospheric heating, with the rest mainly converted into chemical potential energy plus some small airglow loss. The jump to values of ~50% below 30 km altitude is due to the dominant heating from

12

Figure 6. (a) Mean exothermic chemical heating rates QEC for June 1988-2013 at 0.00245 hPa (~90 km altitude) after rebasing all local times to 0000 UTC. Frequency histograms of these heating rates versus temperature at equator at (b) local midnight [0000 LT: 0oE in (a)] and (c) local midday [1200

LT: 180oE in (a)]. Red asterisk shows mean value and blue curve shows linear fit to temperature dependence. (d) mean heating rates 𝑸𝟎 and the modified mean 𝑸�𝟎 given by eq. (4), (e) derivative

term 𝑸𝑻𝟎 and (f) mean temperature term 𝑻𝟎.at the equator, all plotted as a function of hour angle. the Chappuis and long Huggins bands which absorb photons of λ >310 nm and thus produce only O(3P). By ~40 km altitude, higher energy photons absorbed in the Hartley band and Herzberg continuum generate significant O(1D) with resulting increased energy locked in chemical forms D that reduce the residual available for local heating. Likewise, above ~90 km altitude this ratio decreases still further as O2 absorption in the SRC dominates, which, due to the large O2 bond dissociation energy and additional energy locked in O(1D), channels most of the absorbed SRC photon energy into chemical potential energy. Heating Rates due to Exothermic Chemical Reactions Our heating rate parameterization has highlighted the importance of the energy D locked in chemical forms after the primary photolysis reactions. Due to low pressures above ~60 km altitude, collisions among molecules and atoms become ineffective in liberating this energy rapidly as local heating, so that rates of key exothermic recombination reactions become sufficiently long that energy locked within these photolysis products is not rapidly liberated as local heating. Indeed, in the MLT the stored chemical energy D can often be liberated as heat at much later times (up to months), after the consistuents in question have been transported far from their original formation locations, and often with little or no contrast in chemical heat deposition between night and day. Although these processes can be modeled using a detailed photochemistry module within a global model that initializes, transports and photochemically updates several dozen mutually interacting chemical species relevant to the MLT heat budget (Schmidt et al., 2006; Marsh et al., 2007), the computational cost of this approach for an NWP system is prohibitive. Thus we have devised an entirely new and efficient approach to parameterizing this source of atmospheric heating for NAVGEM, which acts as the exothermic chemical heating module depicted in yellow in Figure 3.

13

First, we performed a 25-year simulation (years 1988-2013) using the Whole Atmosphere Community Climate Model (WACCM) in a configuration extending from the ground to 150 km altitude and containing a detailed parameterization of atmospheric photolysis and multiconstituent chemistry and transport for climate and research applications (Marsh et al. 2007). To retain closer associations with observational reality, these runs were performed in the model’s specified dynamics (SD) mode (SD-WACCM: Sassi et al. 2013), in which simulations were continuously nudged to offline 6-hourly global reanalysis fields from 0-50 km altitude provided by NASA’s Modern-Era Retrospective Reanalysis (MERRA). For each month of the simulation, the model’s exothermic chemical heating rates, UV solar heating rates, temperature, ozone, and oxygen densities were averaged and saved globally.

Figure 7. Values of mean heating 𝑸�𝟎 at 0.0052 hPa on day 152 at three different local times, and model 𝑸𝑬𝑬 given by eq. (3) using model temperature fields T.

Figure 6a shows the mean exothermic chemical heating rates QEC from this simulation at 0.00245 hPa, or ~90 km altitude for June. These mean heating rates were derived by shifting rates at different universal times to a common 0000 UTC, so that local midnight is always at 0oE and local midday at 180oE. The plot reveals substantial differences between chemical and shortwave radiative heating rates

14

in that there is little day-night contrast: indeed, at this particular altitude, the chemical heating rates in Figure 6a are somewhat higher at night than during the day. The reason for this is likely due to the importance at this altitude of the exothermic reaction between O3 and H forming OH and O2, which, due to the much greater O3 densities at night relative to the day at this altitude, occurs more frequently and liberates more energy at night relative to the day. Figures 6b and 6c show frequency histograms of QEC versus model temperatures T at the equator in Jume at 0.00245 hPa for local midnight and local midday, respectively, based on all 25 years of simulated WACCM output. These results show that the heating rates have a strong quasi-linear variation with temperature, fitted in a least-squares sense here using the blue line. The strong temperature dependence of these rates occurs for several reason: a strong temperature dependence to the dominant exothermic chemical reaction rates, and variation of local concentrations with vertical wave-induced advection which, via standard parcel advection considerations, is accompanied by adiabatic changes in parcel density and temperature (see, e.g., Smith et al. 2003). These results inform our parameterization approach. We fit a function of the form

𝑄𝐸𝐸(𝑥,𝑦,𝑝,𝑑) = 𝑄0(𝑥, 𝑦, 𝑝,𝑑) + 𝑄𝑇0(𝑥,𝑦,𝑝, 𝑑)[𝑇(𝑥,𝑦,𝑝,𝑑) − 𝑇0(𝑥,𝑦, 𝑝,𝑑)] (2)

= 𝑄�0(𝑥,𝑦,𝑝,𝑑) + 𝑄𝑇0(𝑥,𝑦,𝑝, 𝑑)�𝑇(𝑥, 𝑦, 𝑝,𝑑) − 𝑇�0(𝑥,𝑦,𝑑)� (3)

where

𝑄�0(𝑥, 𝑦, 𝑝,𝑑) = 𝑄0(𝑥, 𝑦,𝑝,𝑑)+𝑄𝑇0(𝑥,𝑦,𝑝, 𝑑)�𝑇�0(𝑥,𝑦, 𝑑) − 𝑇0(𝑥,𝑦,𝑝, 𝑑)� (4)

Here y is latitude, p is pressure, d is the day of the year (season), and x, which is longitude at 0000 UTC, is used as a proxy for the local time dependence. Q0 and T0 are the mean chemical heating rate and temperature, respectively, shown by the red asterisks in Figures 6b and 6c. QT0 is the mean gradient of QEC with respect to temperature T, given by the slope of the fitted blue curves in Figures 6b and 6c. These three parameters are computed for every month at a series of reference longitudes, latitudes and pressures and stored as lookup tables. Since eq. (2) involves a four-dimensional interpolation of these lookup tables onto the NAVGEM model grid at the current (x,y,p,d), we simplify eq. (2) by computing a zonal-mean temperature lookup table 𝑇�0(𝑥, 𝑦, 𝑑), then using the equivalent form of eq. (3), where model temperatures T are now taken as perturbations with respect to 𝑇�0(𝑥,𝑦,𝑑). The mean chemical heating rate then takes a modified form given by eq. (4). Eqs. (3)-(4) give identical results to eq. (2) but reduce the comptuational expense by requiring interpolation of 2 four-dimensional lookup-table fields and 1 three-dimensional lookup-table field, instead of 3 four-dimensional lookup-table fields.

Our lookup-table based on eqs. (3)-(4) and the fitting approach summarized in Figure 6 has been coded up and is currently being tested with input of global NAVGEM temperature fields 𝑇(𝑥, 𝑦,𝑝,𝑑). Figure 7 shows an example at three different local times. Note in particular how the variable NAVGEM temperature inputs convert the smooth 𝑄�0(𝑥, 𝑦, 𝑝,𝑑) rates on the left into a more geophysically variable 𝑄𝐸𝐸(𝑥,𝑦,𝑝, 𝑑) on the right, that yield chemical heating rates in the model that vary more realistically with NAVGEM temperature fields. These NAVGEM tests continue.

QBO-Modulated Seasonal Teleconnections We achieved major milestones in our Task R2 science with the peer-reviewed publication during 2015 of our initial work with the Navy NOGAPS-ALPHA model. This work described how we (a)

15

generated and sustained for the first time in Navy models a realistic QBO in the equatorial lower stratosphere; (b) how this in turn generated an important new seasonal teleconnection in that model, via coupling between the tropical and extratropical straosphere by the Holton-Tan mechanism, and subsequent impacts of Arctic surface late-winter weather through descending NAM anomalies (McCormack et al. 2015). A few examples from this research are given in Figure 8. Figure 8a plots zonal-mean temperature differences from a 10-year model run with an internally generated QBO, driven primarily by a stochastic nonorographic gravity-wave drag parameterizations described in Eckermann (2011). The results here have been separated between westerly and easterly phases of the QBO (QBOW-QBOE). The difference plots highlights the QBO temperature structure in the equatorial stratosphere, but also reveals substantial QBO-modulated temperature differences in the polar winter (northern) hemisphere, leading to a colder stratosphere and warmer mesosphere. This feature is a realistic (observed) response of the so-called Holton-Tan Oscillation, in which QBO winds affect the refractive indices for extratropical planetary-wave propagation, leading to a more stable and colder winter polar stratosphere during QBOW relative to QBOE due to less dynamical disturbances and wave-induced warming circulations in the winter stratosphee due to planetary-wave breaking. This modulation in the strength of the winter stratospheric polar vortex projects strongly onto the NAM and is known to drive stratospheric NAM anomalies that descend over time (Baldwin et al. 2003). Figure 8b shows differences in late winter (March) mean sea-level pressure from two model runs, one without a QBO in the equatorial stratosphere and one with an internally generated QBO. The results show that these QBO modulations in descending NAM anomalies reach the surface, appearing in a statistically significant way in the model’s surface weather patterns during late winter over the Arctic cap. In short, Figure 8b reveals that the model now captures for the first time an important and realistic seasonal teleconnection between the strength/phase of the QBO in the equatorial stratosphere and the strength of the surface NAM affecting Arctic winter weather regimes (see McCormack et al. 2015 for additional details and results).

We have begun work on building a similar capability into the full NAVGEM system, which involves some additional technical issues. For example, whereas NOGAPS-ALPHA uses an Eulerian spectral dynamical core, NAVGEM uses a SISL core. Recent idealized modeling experiments by Yao and Jablonowski (2015) suggest that different dynamical cores can respond in different ways to the internal and parameterized atmospheric wave forcing giving rise to long-period zonal-mean tropical circulations such as the QBO. Likewise, McCormack et al. (2015) showed that levels of vertical and horizontal numerical diffusion in the equatorial lower stratosphere had to be substantially reduced to maintain sharp slowly-descending QBO shear zones. The numerical diffusion characteristics of the NAVGEM dynamical core are different in several ways, with new sources of numerical dissipation such as the decentering damping discussed earlier in this report. A series of runs must therefore be performed to assess the ability of the NAVGEM SISL model to sustain wave-driven QBO circulations and, if necessary, adjusting the model’s diffusion characteristics in the equatorial lower stratosphere until a realistic prognostic QBO arises.

16

Figure 8. (a) Mean differences in zonal-mean zonal temperature between westerly and easterly QBO

phases (QBOW-QBOE) averaged over the December–March period from a 10-year NOGAPS-ALPHA nature run. (b) Differences in ensemble-mean sea-level pressure for March from ensembles

of 120-day runs with QBO activated versus deactivated. The contour interval is 1 hPa. Green contours enclose regions where ensemble mean differences are statistically significant at the 90%

level. See text and McCormack et al. (2015) for more details.

Figure 9 shows some initial NAVGEM experiments in which we have investigated the role of model vertical resolution on the potential simulation of the QBO, based on our earlier work described in McCormack et al. (2015) which indicates that a vertical resolution of ~500m or less was needed to support a realistic QBO in the NOGAPS-ALPHA model. Thus Figure 9a shows a new L80 formulation that, relative to the current L60/L74 formulations in NAVGEM, incorporates substantially improved vertical resolution in the upper troposphere and stratosphere. An initial T119L80 NAVGEM experiment in Figure 9c reveals a persistence of QBO westerlies, but in this run, little descent over time, for reasons we do not as yet understand, requiring further research. For comparison, the T425L60 NAVGEM run in Figure 9d shows a weakening of the initialized QBO westerlies to the uniform easterlies typical of a model that is unable to drive and sustain a realistic QBO. Also evident in this run is unrealistically rapid descent of shear zones.

17

Figure 9. (a) The current L60/L74 vertical level formulations for NAVGEM, plus a new L80 formulation with substantially improved vertical resolution in the upper troposphere and

stratosphere. Experiments were performed for a QBO period where NASA MERRA analysis in (b) shows the observed structure of the QBO. A NAVGEM T119L80 run in (c) shows a realistic persistence of QBO-like westerlies (red) but little descent. An existing T425L60 run shows

weakening to a uniform easterly bias typical of models unable to capture a QBO, and excessively fast shear-zone descent.

IMPACT/APPLICATIONS The Oceanographer of the Navy (Naval Meteorology and Oceanography Command: CNMOC) has selected NAVGEM to be the Navy’s bridging technology from its existing NWP capability at the time (NOGAPS) to a future Earth System Prediction Capability (ESPC) predicting the atmosphere seamlessly across time scales from days to decades. The improved NAVGEM seasonal prediction capabilities provided by this project are moving the Navy closer to a future operational ESPC generating ensemble forecasts on a continuum of time scales spanning weather and climate. Specifically, improved models of deep stratosphere-troposphere coupling provided by this research are improving NAVGEM’s ability to predict deeply coupled regional weather events and systems that are highly relevant for seasonal prediction problems of direct relevance to Navy operations. The salient phenomena, such as Arctic sea ice movement and extent, severe Arctic temperature anomalies, gale-force oceanic winds and high ocean-wave conditions, are all potentially affected, possibly majorly, by deep vertical coupling of NAM/SAM anomalies from the stratosphere to the surface (e.g., Thompson et al. 2002). The research work performed under this project is directly targeting improvements in NAVGEM that will allow it to model and predict these features more accurately and out to longer lead times (weeks to months).

18

TRANSITIONS Key technical aspects of the completed R1 tasks, reflected as new NAVGEM model capabilities such as improved water vapor chemistry and quality control (Task R1c), new hybrid-coordinate vertical layers (Task R1a), and a stochastic parameterization of nonorographic gravity-wave drag (Task R1d), have all transitioned to operations during 2015 as part of the NAVGEM 1.3 transition, as documented in detail in the NAVGEM 1.3 VTR of Whitcomb et al. (2015). In response to this progress, we were also asked to deliver (via telecon) an invited presentation entitled “Influence of Upper Atmospheric Processes” to the National Academy of Sciences (NAS) Committee on Developing a U.S. Research Agenda to Advance Subseasonal to Seasonal Forecasting, convened in Irvine CA, on 21 January 2015. RELATED PROJECTS Doyle, J. D., and S. D. Eckermann, The Boundary Paradox, NRL 6.1 Accelerated Research Initiative,

1 October 2010-30 September 2015.

Eckermann, S. D, and collaborators, New Approaches to the Parameterization of Gravity-Wave and Flow-Blocking Drag due to Unresolved Mesoscale Orography Guided by Mesoscale Model Predictability Research, ONR Award Number: N0001411WX21220, 1 October 2011-30 September 2014.

Hoppel, K. W., High Altitude Data Assimilation for NWP, NRL 6.2 Work Unit, 1 October 2009-September 2013.

REFERENCES Apruzese, J. P., M. R. Schoeberl, and D. F. Strobel (1982), Parameterization of IR cooling in a middle

atmosphere dynamics model: 1. effects on the zonally averaged circulation, J. Geophys. Res., 87(C11), 8951–8966, doi:10.1029/JC087iC11p08951.

Atkinson, R., D. L. Baulch, R. A. Cox, J. N. Crowley, R. F. Hampson, R. G. Hynes,M. E. Jenkin, M. J. Rossi, and J. Troe (2004), Evaluated kinetic and photochemical data for atmospheric chemistry: Volume I – gas phase reactions of Ox, HOx, NOx and SOx species, Atmos. Chem. Phys., 4, 1461–1738, doi:10.5194/acp-4-1461-2004.

Baldwin, M. P., and T. J. Dunkerton (2001), Stratospheric harbingers of anomalous weather regimes, Science, 244, 581–584.

Baldwin, M.P., D.B. Stephenson, D.W.J. Thompson, T.J. Dunkerton, A.J. Charlton, A. O'Neill (2003), Stratospheric memory and extended-range weather forecasts, Science, 301, 636-640.

Bell, C. J., L. J. Gray, A. J. Charlton-Perez, M. M. Joshi, and A. A. Scaife (2009), Stratospheric communication of El Niño teleconnections to European winter. J. Climate, 22, 4083–4096.

Chabrillat, S., and G. Kockarts (1997), Simple parameterization of the absorption of the solar Lyman-alpha line, Geophys. Res. Lett., 24(21), 2659–2662, doi:10.1029/97GL52690, correction, see Chabrillat and Kockarts (1998).

Chabrillat, S., and G. Kockarts (1998), Correction to “simple parameterization of the absorption of the solar Lyman-alpha line”, Geophys. Res. Lett., 25(1), 79, doi:10.1029/97GL03569.

19

Clough, S. A., M. W. Shephard, E. J. Mlawer, J. S. Delamere, M. J. Iacono, K. Cady-Pereira, S. Boukabara, and P. Brown (2005), Atmospheric radiative transfer modeling: a summary of the AER codes, J. Quant. Spectrosc. Radiat. Transfer, 91(2), 233–244, doi:10.1016/j.jqsrt.2004.05.058.

Cordero, E., and A. Staniforth (2004), A problem with the Robert-Asselin time filter for three-time-level semi-implicit semi-Lagrangian discretizations, Mon. Wea. Rev., 132, 600-610.

Coy, L., S. D. Eckermann, K. W. Hoppel, and F. Sassi (2011), Mesospheric precursors to the major stratospheric sudden warming of 2009: Validation and dynamical attribution using a ground-to-edge-of-space data assimilation system, J. Adv. Model. Earth Syst., 3, M10002, 7pp., doi:10.1029/2011MS000067.

Eckermann, S. D. (2011), Explicitly stochastic parameterization of nonorographic gravity-wave drag, J. Atmos. Sci., 68, 1749-1765.

Eckermann, S. D., D. Broutman and H. Knight (2015), Effects of horizontal geometrical spreading on the parameterization of orographic gravity-wave drag. Part 2: Analytical solutions, J. Atmos. Sci., 72, 2348-2365, doi:10.1175/JAS-D-14-0148.1.

Hurrell, J.W., Y. Kushnir, G. Ottersen, and M. Visbeck (2003), The North Atlantic Oscillation: Climate Significance and Environmental Impact, AGU Geographical Monograph Ser. 134, 279pp.

Ineson, S., and A. A. Scaife (2009), The role of the stratosphere in the European climate response to El Niño, Nat. Geosci., 2, 32–36.

Lacoursiére, J., S. A. Meyer, G. W. Faris, T. G. Slanger, B. R. Lewis, and S. T. Gibson (1999), The O(1D) yield from O2 photodissociation near H Lyman-α (121.6 nm), J. Chem. Phys., 110(4), 1949–1958.

Lambert, H.M., A. A. Dixit, E.W. Davis, and P. L. Houston (2004), Quantum yields for product formation in the 120–133 nm photodissociation of O2, J. Chem. Phys., 121(21), 10,437–10,446, doi:10.1063/1.1809114.

Lee, L. C., T. G. Slanger, G. Black, and R. L. Sharpless (1977), Quantum yields for the production of O(1D) from photodissociation of O2 at 1160–1770 Å, J. Chem. Phys., 67(12), 5602–5606, doi:10.1063/1.434759.

Lindzen, R. S., and D. I. Will (1973), An analytic formula for heating due to ozone absorption, J. Atmos. Sci., 30, 513-515.

Marsh, D. R., R. R. Garcia, D. E. Kinnison, B. A. Boville, F. Sassi, S. C. Solomon, and K. Matthes (2007), Modeling the whole atmosphere response to solar cycle changes in radiative and geomagnetic forcing, J. Geophys. Res., 112, D23306, doi:10.1029/2006JD008306.

McCormack, J. P., S. D. Eckermann, and T. F. Hogan (2015), Generation of a quasi-biennial oscillation in an NWP model using a stochastic gravity wave drag parameterization, Mon. Wea. Rev., 43, 2121-2147, doi:10.1175/MWR-D-14-00208.1.

NAS (2010), Assessment of intraseasonal and interannual climate prediction and variability, Committee on Assessment of Intraseasonal to Interannual Climate Prediction and Predictability, National Research Council, The National Academies Press, 192pp (http://www.nap.edu/catalog.php?record_id=12878).

20

Nee, J. B., and P. C. Lee (1997), Detection of O(1D) produced in the photodissociation of O2 in the Schumann-Runge continuum, J. Chem. Phys., 101(36), 6653–6657.

Nissen, K. M., K. Matthes, U. Langematz, and B. Mayer (2007), Towards a better representation of the solar cycle in general circulation models, Atmos. Chem. Phys., 7(20), 5391–5400.

Remsberg, E. E., B. T. Marshall, M. Garcia-Comas, D. A. Krueger, G. S. Lingenfelser, F. J. Martin-Torres, M. G. Mlynczak, J. M. Russell III, A. K. Smith, Y. Zhao, C. W. Brown, L. L. Gordley, M. López-Gonzalez, M. López-Puertas, C.-Y. She, M. J. Taylor, and R. E. Thompson (2008), Assessment of the quality of the version 1.07 temperature-versus-pressure profiles of the middle atmosphere from TIMED/SABER, J. Geophys. Res., 113(D17), D17101, doi:10.1029/2008JD010013.

Rivest, C., A. Staniforth, and A. Robert (1994), Spurious resonant response of semi-Lagrangian discretization to orographic forcing: Diagnosis and solution. Mon. Wea. Rev., 122, 366-376.

Sassi, F., H.-L. Liu, J. Ma, and R. R. Garcia (2013), The lower thermosphere during the northern hemisphere winter of 2009: a modeling study using high-altitude data assimilation products in WACCM-X, J. Geophys. Res., 118, 8954-8968, doi:10.1002/jgrd.50632.

Schmidt, H., G. P. Brasseur, M. Charron, E. Manzini, M. A. Giorgetta, T. Diehl, V. I. Fomichev, D. Kinnison, D. Marsh, and S. Walters (2006), The HAMMONIA chemistry climate model: sensitivity of the mesopause region to the 11-year solar cycle and CO2 doubling, J. Climate, 19, 3903–3931.

Smith, A. K., D. R. Marsh, and A. C. Szymczak (2003), Interaction of chemical heating and the diurnal tide in the mesosphere, J. Geophys. Res., 108(D5), 4164, doi:10.1029/2002JD002664.

Strobel, D. F. (1978), Parameterization of the atmospheric heating rate from 15 to 120 km due to O2 and O3 absorption of solar radiation, J. Geophys. Res., 83(C12), 6225–6230, doi:10.1029/JC083iC12p06225.

Sukhodolov, T., E. Rozanov, A. I. Shapiro, J. Anet, C. Cagnazzo, T. Peter, and W. Schmutz (2014), Evaluation of the ECHAM family radiation codes performance in the representation of the solar signal, 7(6), 2859–2866, doi:10.5194/gmd-7-2859-2014.

Thompson, D. W. J., M. P. Baldwin, and J. M. Wallace (2002), Stratospheric connection to Northern Hemisphere wintertime weather: Implications for prediction, J. Climate, 15, 1421-1428.

Tripathi, O. P., M. Baldwin, A. Charlton-Perez, M. Charron, S. D. Eckermann, E. Gerber, R. G. Harrison, D. R. Jackson, B.-M. Kim, Y. Kuroda, A. Lang, S. Mahmood, R. Mizuta, G. Roff, M. Sigmond, and S.-W. Son (2015a), The predictability of the extratropical stratosphere on monthly time-scales and its impact on the skill of tropospheric forecasts, Quart. J. Roy. Meteor. Soc., 141, 987-1003, doi:10.1002/qj.2432.

Tripathi, O. P., M. Baldwin, A. Charlton-Perez, M. Charron, J. C. H. Cheung, S. D. Eckermann, E. Gerber, D. R. Jackson, Y. Kuroda, A. Lang, J. Mclay, R. Mizuta, C. Reynolds, G. Roff, M. Sigmond, S.-W. Son, and T. Stockdale (2015b), Examining the predictability of the stratospheric sudden warming of January 2013 using multiple NWP systems, Mon. Wea. Rev., submitted.

WCRP (2008), Position paper on seasonal prediction, First WCRP Seasonal Prediction Workshop, 4-7 June 2007, Barcelona, Spain, Report No. 3/2008, ICPO Publication No. 127, 23pp.

(http://www.wcrp-climate.org/documents/WCRP_SeasonalPrediction_PositionPaper_Feb2008.pdf).

21

Weare, B. C., C. Cagnazzo, P. G. Fogli, E. Manzini, and A. Navarra (2012), Madden-Julian Oscillation in a climate model with a well-resolved stratosphere, J. Geophys. Res., 117, D01103, doi:10.1029/2011JD016247.

Whitcomb, T., W. Clune, T. Hogan, R. Pauley, B. Ruston, C. Skupniewicz, S. Eckermann, T. Maurer, J. McCormack, E. J. Metzger, J. Moskaitis, G. Poe, J. Ridout, P. Shen, S. Swadley, K. Viner, and A. Ulina (2015), Validation Test Report for the Navy Global Environmental Model (NAVGEM) 1.3 System, Navy Atmospheric Models Oversight Panel, 65 pp, 3 March 2015.

Yao, W., and C. Jablonowski (2015), Idealized quasi-biennial oscillations in an ensemble of dry GCM dynamical cores. J. Atmos. Sci., 72, 2201–2226, doi: http://dx.doi.org/10.1175/JAS-D-14-0236.1.

Zhu, X. (1994), An accurate and efficient radiation algorithm for middle atmosphere models, J. Atmos. Sci., 51, 3593–3614.

Zhu, X., J. H. Yee, and E. R. Talaat (2003), Effect of short–term solar ultraviolet flux variability in a coupled model of photochemistry and dynamics, J. Atmos. Sci., 60, 491–509.

PUBLICATIONS Coy, L., and C. A. Reynolds (2014), Singular vectors and their nonlinear evolution during the January

2009 stratospheric sudden warming, Quart. J. Roy. Meteorol. Soc., 140, 1013-1024, doi: 10.1002/qj.2181.

Eckermann, S. D. and J. P. McCormack (2015), Extension of linearized parameterizations of gas-phase stratospheric ozone photochemistry through the upper stratosphere, mesosphere and lower thermosphere, (paper in preparation).

Eckermann, S. D., J. P. McCormack, J. Ma, T. F. Hogan, and K. A. Zawdie (2014), Stratospheric analysis and forecast errors using hybrid and sigma coordinates, Mon. Wea. Rev., 142, 476-485.

Eckermann, S. D., D. Broutman and H. Knight (2015), Effects of horizontal geometrical spreading on the parameterization of orographic gravity-wave drag. Part 2: Analytical solutions, J. Atmos. Sci., 72, 2348-2365, doi:10.1175/JAS-D-14-0148.1.

Hogan, T. F., M. Liu, J. A. Ridout, M. S. Peng, T. R. Whitcomb, B. C. Ruston, C. A. Reynolds, S. D. Eckermann, J. R. Moskaitis, N. L. Baker, J. P. McCormack, K. C. Viner, J. G. McLay, M. K. Flatau, L. Xu, C. Chen, and S. W. Chang (2014), The Navy Global Environmental Model, Oceanography, 27(3), 116-125, doi:10.5670/oceanog.2014.73.

Hogan, T., M. Peng, N. Baker, C. Reynolds, B. Ruston, M. Liu, J. Ridout, S. Eckermann, J. Moskaitis, T. Whitcomb, K. Viner, J. McLay, P. Pauley, L. Xu., R. Langland, M. Flatau, J. McCormack, and S. Chang (2014), The Navy Global Environmental Model, NRL Review 2013, NRL/PU/3430--14-577, RN:14-1231-1938, June 2014, pp 138-140.

Hoppel, K. W., S. D. Eckermann, L. Coy, G. E. Nedoluha, D. R. Allen, S. D. Swadley, and N. L. Baker (2013), Evaluation of SSMIS upper atmosphere sounding channels for high-altitude data assimilation, Mon. Wea. Rev., 141, 3314-3330.

McCormack, J. P., S. D. Eckermann, and T. F. Hogan (2015), Generation of a quasi-biennial oscillation in an NWP model using a stochastic gravity wave drag parameterization, Mon. Wea. Rev., 43, 2121-2147, doi:10.1175/MWR-D-14-00208.1.

22

Tripathi, O. P., M. Baldwin, A. Charlton-Perez, M. Charron, S. D. Eckermann, E. Gerber, R. G. Harrison, D. R. Jackson, B.-M. Kim, Y. Kuroda, A. Lang, S. Mahmood, R. Mizuta, G. Roff, M. Sigmond, and S.-W. Son (2015a), The predictability of the extratropical stratosphere on monthly time-scales and its impact on the skill of tropospheric forecasts, Quart. J. Roy. Meteor. Soc., 141, 987-1003, doi:10.1002/qj.2432.

Tripathi, O. P., M. Baldwin, A. Charlton-Perez, M. Charron, J. C. H. Cheung, S. D. Eckermann, E. Gerber, D. R. Jackson, Y. Kuroda, A. Lang, J. Mclay, R. Mizuta, C. Reynolds, G. Roff, M. Sigmond, S.-W. Son, and T. Stockdale (2015b), Examining the predictability of the stratospheric sudden warming of January 2013 using multiple NWP systems, Mon. Wea. Rev., submitted.

Whitcomb, T., W. Clune, T. Hogan, R. Pauley, B. Ruston, C. Skupniewicz, S. Eckermann, T. Maurer, J. McCormack, E. J. Metzger, J. Moskaitis, G. Poe, J. Ridout, P. Shen, S. Swadley, K. Viner, and A. Ulina (2015), Validation Test Report for the Navy Global Environmental Model (NAVGEM) 1.3 System, Navy Atmospheric Models Oversight Panel, 65 pp, 3 March 2015.

HONORS/AWARDS/PRIZES 1. Stephen Eckermann received the 2013 Sigma-Xi Award for Applied Science from the NRL

Edison Chapter for foundational contributions to the Navy’s high-altitude forecasting and analysis capabilities.

2. Stephen Eckermann was awarded the 2013 American Meteorological Society’s Editor’s Award of the Journal of the Atmospheric Sciences for “thorough and detailed reviews with well-informed, well-posed and carefully argued questions for authors”: see http://www.ametsoc.org/awards/2013awardrecipients.pdf

3. Stephen Eckermann was one of ten NRL awardees of the Navy Acquisition Excellence Award for Technology Transition, for the Navy Global Environmental Model (NAVGEM): see http://www.nrl.navy.mil/media/news-releases/2013/nrl-receives-navy-acquisition-excellence-award-for-global-weather-prediction-model

4. Stephen Eckermann delivered (via telecon) an invited presentation entitled “Influence of Upper Atmospheric Processes” to the National Academy of Sciences (NAS) Committee on Developing a U.S. Research Agenda to Advance Subseasonal to Seasonal Forecasting, convened in Irvine CA, 21 January 2015.