Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events: a numerical study with WRF high-resolution simulations

Earth Syst. Dynam., 10, 485–499, 2019 https://doi.org/10.5194/esd-10-485-2019 © Author(s) 2019. This work is distributed under the Creative Commons Attribution 4.0 License. Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events: a numerical study with WRF high-resolution simulations Miguel A. Prósper1, Ian Sosa Tinoco2, Carlos Otero-Casal1,3, and Gonzalo Miguez-Macho1 1Universidade de Santiago de Compostela, Galicia, Spain 2Technical Institute of Sonora, Sonora, Mexico 3MeteoGalicia, Xunta de Galicia, Santiago de Compostela, Galicia, Spain Correspondence: Miguel A. Prósper (prosper[email protected]) Received: 26 January 2019 – Discussion started: 18 February 2019 Revised: 25 June 2019 – Accepted: 6 July 2019 – Published: 30 July 2019 Abstract. Tehuantepecers or Tehuanos are extreme winds produced in the Isthmus of Tehuantepec, blowing south through Chivela Pass, the mountain gap across the isthmus, from the Gulf of Mexico into the Pacific Ocean. They are the result of the complex interaction between large-scale meteorological conditions and local orographic forcings around Chivela Pass, and occur mainly in winter months due to cold air damming in the wake of cold fronts that reach as far south as the Bay of Campeche. Even though the name refers mostly to the intense mountain gap outflow, Tehuantepecer episodes can also generate other localized extreme wind conditions across the region, such as downslope windstorms and hydraulic jumps, which are strong turbulent flows that have a direct effect on the Pacific side of the isthmus and the Gulf of Tehuantepec. This study focuses on investigating these phenomena using high horizontal and vertical resolution WRF (Weather Research and Forecasting) model simulations. In particular, we employ a 4-nested grid configuration with up to 444 m horizontal spacing in the innermost domain and 70 hybrid-sigma vertical levels, 8 of which lie within the first 200m above ground. We select one 36 h period in December 2013, when favorable conditions for a strong gap wind situation were observed. The high-resolution WRF experiment reveals a significant fine-scale structure in the strong Tehuano wind flow, beyond the well known gap jet. Depending on the Froude number upstream of the topographic barrier, different downslope windstorm conditions and hydraulic jumps with rotor circulations develop simultaneously at different locations east of Chivela Pass with varied crest height. A comparison with observations suggests that the model accurately represents the spatially heterogeneous intense downslope windstorm and the formation of mountain wave clouds for several hours, with low errors in wind speed, wind direction and temperature. 1 Introduction The Isthmus of Tehuantepec, in the Mexican state of Oaxaca, is the narrowest stretch of land separating the Gulf of Mexico from the Pacific Ocean. The Sierra Madre mountains cross the isthmus from east to west, but leaving a pronounced gap in the middle (Chivela Pass), coinciding with the point of shortest distance between the two sea masses of only 200km. The elevation of the Chivela Pass is 224 m, whereas mountain peaks in the side sierras reach 2000 m, creating ideal conditions to generate a powerful wind corridor (Romero-Centeno et al., 2003). In winter, cold high-pressure systems originating in North America move over the Gulf of Mexico in the wake of south-reaching cold fronts, and large pressure differences develop across the isthmus between the Bay of Campeche and the Gulf of Tehuantepec, on the Pacific side. This pressure gradient results in a northerly wind situation in which the flow is accelerated southward by cold air damming, traveling through Chivela Pass to finally blow violently outward into the Pacific Ocean as it fans out and Published by Copernicus Publications on behalf of the European Geosciences Union. 486 M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events curves anticyclonically. This gap outflow extends for hundreds of kilometers over the Gulf of Tehuantepec, a result of the combined effect of the open marine surface and the low Coriolis parameter of the tropical latitude of the area. The small impact of planetary rotation yields very weak synoptic forcing; thus, the gap wind does not encounter any ambient large-scale flow to merge with, describing a very-close-toinertial trajectory with a large radius, also as a consequence of the diminished Coriolis force (Steenburgh et al., 1998). In the Gulf of Tehuantepec, the strong sea surface wind stress generates intense upwelling and vertical mixing in the upper ocean (Hong et al., 2018). The largest number of these events tends to occur in December, with a mean duration of 48 h (Brennan et al., 2010). These powerful mountain gap winds are called Tehuantepecers or Tehuanos and have been the focus of several previous studies (Brennan et al., 2010; McCreary et al., 1989; Jaramillo and Borja, 2004) detailing their general setting, drivers and dynamics (Steenburgh et al., 1998) of strong wind situations in the isthmus. Little knowledge exists, however, on the fine-scale structure of the Tehuantepecer flow elsewhere across the isthmus, away from Chivela Pass. There is evidence from observations and earlier numerical studies (Steenburgh et al., 1998) that the low-elevation topography of Chivela Pass can excite mountain waves and also that these are not only restricted to the pass itself but extend into the much higher mountain crests to the west and especially to the east, as the cold air pool is often thick enough to surpass them. This gravity wave activity can potentially result in downslope windstorms (DSWSs hereafter) producing severe turbulent phenomena such as rotors and hydraulic jumps (HJs hereafter) (Durran, 1986a; Sheridan and Vosper, 2006) on the Pacific side of the isthmus, to the lee of local orography, which would explain several accidents related to strong winds reported by the Oaxaca’s Civil Protection Commission (Santiago, 2018; Hernández, 2018; Televisa, 2018; Rodríguez, 2018) every year during some Tehuano occurrences. The ability to understand and forecast these events is very relevant since the isthmus has been an important development site for wind farms since the 2000s (Coldwell et al., 2017). Currently this region allocates 76.8% of the wind power capacity installed in Mexico, with approximately 2360 MW (Baxter et al., 2017), which is expected to double to 5076 MW by 2020 (Francisco Ibáñez, 2018). The main goal of the present work is to study the variability in flow behavior, beyond the well-known mountain gap jet, in Tehuano wind episodes across the Isthmus of Tehuantepec, depending on topographic barrier height and thermodynamic conditions of the air mass, using highresolution simulations with the Weather Research and Forecasting (WRF) model. Many studies have successfully employed WRF to analyze these kinds of mountain-flow events in other parts of the world. In the US, for example, Pokharel et al. (2017b) study a DSWS and HJ to the lee of the Medicine Bow Mountains in southeast Wyoming. Another DSWS in this same area is also investigated by Grubiši´ c et al. (2015) with WRF. Other studies such as Cao and Fovell (2016), Pokharel et al. (2017a), Prtenjak and Belusic (2009), Jung-Hoon and Chung (2006), Ágústsson and Ólafsson (2014), and Priestley et al. (2017) also use WRF at high resolution to analyze mountain wave flows in different locations. To the best of our knowledge, there is no previous work that studies in detail lee wave phenomena in the Mexican state of Oaxaca. The high-resolution WRF simulations employed in the present study allow us to obtain a more complete knowledge of these events, from the synoptic scale to the small scale, focusing on the downslope winds and the HJ that develop along the mountain ranges neighboring Chivela Pass. The article is organized as follows: in Sect. 2 the methodology is explained in detail, from the climatology of the region to the model configuration. In Sect. 3, the primary results obtained are shown, divided into synoptic-mesoscale situations and the upstream–downstream structure of the phenomena. Finally, in Sect. 4 the conclusions reached are discussed. 1.1 Mountain wave phenomena, hydraulic analog and relation with the Froude number DSWSs result from the intense flow acceleration occurring on the lee slope of a mountain under certain cross-barrier wind conditions. They resemble the behavior of water flow in an open channel when encountering an obstacle, such as when a relatively slow river increases its speed when flowing in a thin layer over a mill’s dam or over a rock. As in the case of the river, DSWSs often end abruptly with a return to the state upstream of the obstacle through a turbulent HJ somewhere downstream. The similarity in both fluid’s behavior suggests that the physical processes behind them are also alike, and that shallow-water theory could be applied somehow to the atmosphere (Long, 1953). However, the complexity of the unbounded atmospheric flow, without a free surface, makes it difficult to make the analog so simple because gravity waves in the atmosphere propagate vertically in addition to horizontally as in shallow water, and nonlinear effects are important. Considerable observational and numerical experiments have been developed to elucidate the dynamics of atmospheric mountain lee flows (see Durran, 1990, for a review). The current view is that in the atmosphere DSWSs are observed when stable air at low levels flows, similarly to shallow water, as if having a free surface somewhere above the obstacle preventing vertical energy dissipation. This occurs when gravity waves do not propagate vertically and break because of the presence of a critical level (Durran and Klemp, 1987) or due to overturning for having an amplitude that is too large (Clark and Peltier, 1977), or when without wave breaking there is an interface separating highly stable lower layers from less-stable air above (Durran, 1986b; Klemp and Durran, 1987; Bacmeister and Pierrehumbert, 1988). Wave breaking creates a well-mixed layer to the lee of the obstaEarth Syst. Dynam., 10, 485–499, 2019 www.earth-syst-dynam.net/10/485/2019/ M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events 487 cle, which generates a dividing streamline separating undisturbed flow aloft and trapped energy and flow analogous to hydraulics in the lower surficial branch (Smith, 1985b). A highly stable lower layer topped by less-stable air results in reflecting and decaying waves aloft, enhanced nonlinear effects, and the atmosphere also flowing like a two-layer fluid and behaving like shallow water when encountering the barrier. In shallow water theory, the Froude number (Fr), which is the ratio of the mean speed to the intrinsic gravity wave phase speed in the fluid, determines its behavior when encountering the obstacle, depending on whether gravity is balanced mostly by acceleration (F r > 1) or pressure gradient forces (F r < 1). HJs and significant flow acceleration to the lee of the obstacle occur when the fluid transitions from a subcritical (F r < 1) to supercritical (F r > 1) regime at the top of the barrier. It is not straightforward to define a Froude number to determine the regime of atmospheric flows (Sheridan and Vosper, 2006; Smith, 1989, 1985a) as it is for shallow water because there is no clear analog to flow depth controlling gravity wave speed and pressure perturbations. Key factors determining wave properties and flow characteristics in this case are stratification, wind speed and barrier height, or rather the relative values among them. A quantity that combines these three variables, and is referred to as Froude number for shallow atmospheric flows over mountains by many authors, is the following (Smith, 1989): F r =U NH ,(1) where Uis the flow speed, Nthe Brunt–Väisäla frequency (Eq. 2) and Hthe mountain height. N=sg θ0 dθ dz(2) with gbeing the acceleration due to gravity, dθ/dtthe potential temperature gradient in the stable layer and θ0the potential temperature at the base of this layer. To avoid confusion with the classical Froude number, the inverse of Fr defined as above is often used instead and called the nondimensional mountain height. Fr is a measure of flow deceleration and stagnation upwind from the mountain (Baines, 1987). It can also be regarded as an estimate of nonlinearity. When F r 1 there are significant nonlinear effects and blocking, whereas for F r 1 the opposite occurs (Smolarkiewicz and Rotunno, 1989). Fr around 1 indicates a transitional regime between the two states and favorable conditions for the formation of DSWS and HJs. 2 Methodology 2.1 WRF configuration We use the Advanced Research WRF (ARW) model (Skamarock et al., 2008) version 3.9 (WRFV3.9) to perform the simulations. Based on a fully compressible and nonhydrostatic dynamic core, WRFV3.9 is a limited-area mesoscale and microscale model, with a terrain-following hydrostaticpressure vertical coordinate designed for operational forecasting, as well as research. For the experiments, we employ a nested domain configuration in order to achieve sufficiently high resolution in the innermost grids to capture the smallscale structure of the flow, while reproducing the synoptic phenomenology conducive to local DSWS in the parent one (Fig. 1a). The domain’s configuration meets the requirements recommended by Warner (2011), including a parent (d01) and four nested grids (d02, d03, d04 and d05) (Fig. 1a) that are one-way interacting. D01 is centered at 17.91◦N, 93.44◦W (Fig. 1a) with 80×80 grid points of 36 km of horizontal resolution. The horizontal resolutions of d02–d05 are 12 km (91 ×91 grid points), 4 km (94 ×94 grid points), 1.3 km (103×103 grid points) and 444 m (106×106 grid points) respectively. D03 covers the whole isthmus area, with Chivela Pass approximately at its center, while the highest resolution domains, d04 and d05, are slightly displaced to the south and east, respectively. D04 includes Chivela Pass and the section of the Sierra Madre de Chiapas range east of it, with heights reaching about 2000 m. For its part, the finest grid D05 encompasses the southernmost hills to the east of Chivela Pass and the coastal plain at their base. This domain configuration focuses the area of interest of the study on the Pacific side of the Tehuano, encompassing the flow path before and after crossing the mountains, the exit section of the mountain gap and the gradually rising mountains east of it, around the only two available observational sites for validation, labeled as MET1 and MET2 in Fig. 1b. All the domains have 70 hybrid-sigma vertical levels, 8 of which lie within the first 200m above ground at about 16, 46, 71, 96, 122, 147, 173 and 198 m heights. The hybrid sigmapressure vertical coordinate follows the terrain near the surface and gradually transitions to constant pressure at higher levels. The benefit of this vertical coordinate system is a numerical noise reduction in the upper layers over mountains (Powers et al., 2017). We maintain this fine vertical grid spacing in all the domains to capture as wide a range of motions as possible over the depth of the boundary layer. Land use information for d04 and d05 is obtained from the ESA CCI (European Space Agency Climate Change Initiative) database (Bontemps et al., 2013) with a resolution of 300 m. The terrain elevation used comes from the ASTER Global Digital Elevation Map (GDEM) from USGS (United States Geological Survey) (Slater et al., 2009) with a resolution of 30m. In the other domains, terrain and land use data are from the WRF global standard database, both at 30 arcsec resolution for d03 and 2 arcmin for d02 and d01. We simulate a 36h period, from 23 December 12:00 to 25 December 2013 00:00 UTC, which registers DSWS conditions in the observational data. Regarding the main physics options, the simulations use the tropical suite configuration www.earth-syst-dynam.net/10/485/2019/ Earth Syst. Dynam., 10, 485–499, 2019 488 M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events Figure 1. WRF nested domain configuration. (a) Coarser three domains with their number of grid points. (b) Higher-resolution domains (d04 and d05), both with their respective topographies (meters above sea level). MET1 and MET2 are the locations of the meteorological stations used as validation points. The two red lines represent the vertical cross sections shown in Figs. 3 and 5. (Table 1), introduced in WRF version 3.9, except for the planetary boundary layer, which is parametrized by the Shin– Hong scale-aware scheme (S–H) (Shin and Hong, 2015). Table 1 summarizes this physics configuration used. The S–H planetary boundary layer option is more suitable for the high resolution of the innermost domain (444 m) because it helps to mitigate a double counting effect of the small-scale processes in gray-zone resolutions. Apart from this, this scheme provides a turbulent kinetic energy (TKE) diagnostic variable useful for our analyses. 2.2 Global model and real data Global Forecast System (GFS) analysis data from the National Centers for Environmental Prediction (NCEP) are used as initial and boundary conditions for the WRF model, with a 3 h update interval. The horizontal resolution of this dataset for all variables is 0.5◦×0.5◦with 32 vertical levels ranging from 1000 to 10 hPa. The observational data used in this work are provided by the Mexican National Laboratory of remote sensors (https://clima.inifap.gob.mx/lnmysr/ Estaciones/MapaEstaciones, last access: 10 January 2018), collected every 15min at two meteorological stations whose locations are presented in Table 2 and marked in Fig. 1b as MET1 and MET2. Wind speed, wind direction and temperature at 3m height from these points are used for validation. Model data are extrapolated from their native sigma levels to the height of the meteorological station using Eq. (3), which relates wind speed with friction wind speed, which is also a diagnostic variable in the model. wsz=U∗ Kln( z z0),(3) where U∗in the friction velocity, Kis the von Kármán constant, zis the height and z0is the surface roughness (m). 3 Results and discussion 3.1 Synoptic and mesoscale situation Figure 2 shows the synoptic situation, from GFS analysis data, in North America and Central America 18h before (Fig. 2a) and at peak intensity of the event (Fig. 2b). The large-scale setting is typical of Tehuantepecer wind episodes where an Arctic air mass east of the Rockies pushes south across the Great Plains with its leading edge reaching first the Gulf of Mexico (on 22 December, Fig. 2a) and then as far south as the Bay of Campeche (about 1d later, Fig. 2b), the result of cold air damming east of the Sierra Madre Oriental range in Mexico. The equatorward displacement of the associated high-pressure system on the wake of the cold front creates a strong pressure gradient across the Isthmus of Tehuantepec, ultimately producing the strong mountain gap winds through the low elevation of Chivela Pass. The general situation favoring Tehuantepecer winds persists for about 6 d, more pronounced earlier on. Figure 3 shows the mesoscale conditions of the fully developed extreme wind episode at 03:00UTC of 24 December 2013 from the WRF simulation. In the coarser grid D01 (36 km spacing, outlined in white in Fig. 2) the cold air damming by the Sierra Madre Oriental is clearly apparent with the northerly cool air mass intrusion extending to the Earth Syst. Dynam., 10, 485–499, 2019 www.earth-syst-dynam.net/10/485/2019/ M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events 489 Table 1. Main physics parameterizations used. Microphysics Hong and Lim (Hong and Lim, 2006) Cumulus Zhang and Wang (Zhang et al., 2011)∗disabled in d04 and d05 Long-wave radiation RRTMG (Iacono et al., 2008) Short-wave radiation RRTMG (Iacono et al., 2008) Planet boundary layer Shin and Hong (Shin and Hong, 2015) Surface layer option Revised MM5 surface layer (Jiménez et al., 2012) Land-surface physics Noah land-surface (Tewari et al., 2004) Table 2. Weather station positions. Name Latitude Longitude Elevation Height (◦) (◦) (m) (m) MET1 Santiago Niltepec 16.5535 −94.6439 65 3 MET2 Ixhuatan 16.3673 −94.4717 18 3 Figure 2. (a) 850 hPa temperature (◦C) and sea level pressure (hPa) from GFS 0.5 analysis data at 22 December 2013 18:00UTC. (b) Same as (a) at 24 December 2013 03:00 UTC. D01 simulation domain is represented with a white square in both cases. Bay of Campeche, from where it is funneled across the Isthmus of Tehuantepec (Fig. 3a). The higher resolution of the nested grids shows in greater detail the structure of the gap winds. Figure 3b, from the first nested D02 (12km grid spacing), depicts the wind field (vectors) highlighting in shades the values above 10 m s−1at about 70 m above the surface (sigma level 3), the approximate height of wind turbine hubs. The strong Tehuano outflow from Chivela Pass reaches velocities of 25 m s−1as it fans out for more than 100 km into the Pacific Ocean. High wind speeds are not only restricted to the outflow of the mountain gap itself but the simulation results suggest that they also occur in the mountains west and especially east of Chivela Pass. The potential temperature field on the same σ=3 level at the enhanced resolution (4 km) of the next nested grid (D03; Fig. 3c) illustrates how the stable cold air mass to the north surmounts the lower hills neighboring the pass, particularly those to the east where elevation increases more gradually. As shown in the following Sect. 3.2, acceleration on the top of these mountains and to their lee is related to gravity wave activity, which results in the development of strong downslope winds, rotors and HJs. The model sounding at location NP, on the Gulf of Mexico side of the isthmus (Fig. 3d), shows in the temperature profile a stable lower boundary layer capped by a very stable isothermal layer from 850 hPa up to about 2000 to 2500m, defining the depth of the cold air pool, indeed above the aforementioned mountain tops. A weak inversion associated with the subsidence within the high-pressure system aloft is also www.earth-syst-dynam.net/10/485/2019/ Earth Syst. Dynam., 10, 485–499, 2019 490 M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events Figure 3. 24 December 2013 03:00 UTC (a) 850hPa temperature (◦C), sea level pressure (hPa) and wind vectors in the parent grid d01, (b) wind speed (values >10 m s−1) and wind vectors at σ=3 (about 70m above ground) in d02, and (c) topography (m, contours) and potential temperature (K, shades) and wind vectors at σ=3 in d03. (d) Skew-T log-P diagram on the Gulf of Mexico side of the isthmus, northeast of Chivela Pass (NP point in c; lat =17.6463◦, long = −95.6246◦) with wind barb profile at right. clearly apparent just below 500hPa. Above this level, winds are weak and veer from being southeasterly to southwesterly in the upper troposphere. Below 500 hPa, winds back from an easterly to a northeasterly direction at about 800 hPa, and more strongly in lower levels, becoming westerly at the surface, indicating intense cold air advection. There is a pronounced reverse wind shear in the lower troposphere. As mentioned earlier, the only available observations in the area are from stations MET1 and MET2, whose positions are marked in Fig. 3c. They are both located on the Pacific coastal plain, south of the mountains bordering Chivela Pass to the east; MET1 is closer to the relief and further west than MET2. The wind speed time series covering the entire Tehuano episode for both stations is shown in Fig. 4. Wind speeds are low in the previous days and show a daily cycle, likely linked to sea breezes, more clearly evident in station MET2 closer to the coast. The situation changes after about 23 December 06:00 UTC when, picking up intensity, wind speeds become more constant throughout the day, the signature of a Tehuano wind occurrence. At about 00:00 UTC on Earth Syst. Dynam., 10, 485–499, 2019 www.earth-syst-dynam.net/10/485/2019/ M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events 491 Figure 4. Wind speed (m s−1) time series for observation stations MET1 and MET2 (locations marked in Fig. 3). The orange shaded area corresponds to the period simulated. 29 December, the episode decays and winds go back to local breeze regimes. The extent of the simulated period covers the first 36 h of the event (shaded in Fig. 4), corresponding with its highest intensity in both locations. These observations away from Chivela Pass show evidence, as the simulations suggest, that the neighboring mountains may also induce strong winds, even though they certainly do not reach out as far as the mountain gap outflow does. 3.2 Upstream–downstream structure In this section, we focus on the fine-scale structure of the flow acceleration across the isthmus, and more precisely on that occurring on the westernmost section of the Sierra Madre de Chiapas bordering Chivela Pass, apart from the well-known strong gap wind jet. This is the area covered by the d04 domain of 1.3 km resolution (highlighted in green in Fig. 3c), encompassing with 137 km from north to south the wind flow path before and after crossing the mountains. Figure 5 depicts, from d04, latitudinal cross sections (red lines in Fig. 1b) facing east (south to the left, north to the right) of different variables at the longitude of the two validation points MET1 (Fig. 5a–f) and MET2 (Fig. 5g–l). The tight isentropes on the windward side of the mountains (to the right in Fig. 5a, d, g and j), indicate a quite similar stable stratification in the whole lower tropospheric column in both cases, stronger in the layers below about 2.5 km and weaker above. The temperature profile in Fig. 3d show evidences that the 2500 m height level corresponds to the depth of the cold air. Winds are northerly in these lower layers to the north of the mountains with somewhat higher speeds above 20ms−1upstream from MET1 than further east, north of MET2. In both cases, winds above the cool pool are much weaker, and back to an easterly component at about 4000 m and above. The latter height is therefore a critical level since the cross-mountain flow component becomes null; thus, triggered gravity waves will break and dissipate when they approach it. Markowski and Richardson (2010) outline seven conditions conducive to DSWS, albeit not all of them absolutely necessary. These conditions are (1) an asymmetric mountain with steeper lee than windward side (Fig. 5), (2) crossed by strong winds (>15 m s−1) (Fig. 5), with (3) winds mostly normal to the barrier (Figs. 3 and 5), (4) a stable layer above the top and less stable above that (Figs. 3d and 6c), and with (5) cold air advection and largescale subsidence to maintain the stability (Figs. 3d and 6c). Apart from this, (6) reverse wind shear above (Figs. 3d and 6c) and (7) no cool pool in the lee (Figs. 3a and 6b) are also desirable. These conditions are all perfectly met for both locations analyzed, as discussed previously, and indeed intense DSWSs occur in both cases. The stably stratified cross-barrier flow displays wave activity from early on, and wave breaking, as the vertical isentropes suggest (Fig. 5a and d), enhances turbulent mixing (Fig. 5c and f), and yields a region of weak stability and reverse flow immediately downwind from the mountain crests (Fig. 5g and j). In both cases, isentropes on the windward side sink sharply under these layers of low stability on the lee side, much more pronouncedly for the tallest mountain (Fig. 5j). Encompassing the well-mixed region to the lee, a split streamline develops (Smith, 1985b) and below its lower branch there is flow thinning and a significant increase in wind speed. However, the particular features existent on the lee side differ, depending of the height of the topographic obstacle. The strong accelerated flow confined by layers of strong wind shear and turbulence extends for many kilometers downwind from the lowest mountain (Fig. 5i), while ending with a HJ and rotors at the foot when the barrier is higher (Fig. 5j). The formation of either of these lee wave events is related to the Froude number upstream (Smith, 1989). We note that only the turbulence produced by the existent wind shear in the column is well captured in the simulation (Fig. 5i and l). The subgrid-scale turbulence associated with gravity waves due to rotors and nonlocal turbulent advection or with www.earth-syst-dynam.net/10/485/2019/ Earth Syst. Dynam., 10, 485–499, 2019 492 M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events Figure 5. d04 vertical cross sections (a, d, g, j) of potential temperature (K, contours), wind speed (m s−1shades), and wind barbs referenced to the orientation of the cross section plane, (b, e, h, k) vertical wind component W(m s−1) and (c, f, i, l) wind speed (ms−1) isolines and turbulent kinetic energy (TKE, Jkg−1, shades). Panels (a–c) for MET1 and (d–f) for MET2 correspond to the initial stages of the episode on 23 December 2013 15:00 UTC. Panels (f, h, i) at MET1 and (j, k, l) at MET2 are for the fully developed events on 24 December 2013 03:00 UTC. UP1 and UP2 are the locations where the Froude number upstream the mountain is calculated. instability resulting from wave overturning is not represented and accounted for since a much higher horizontal and vertical resolution on the order of tens of meters would be needed in order to explicitly resolve these features (Vosper et al., 2018). Figure 6a plots the Froude number (Eq. 2) calculated using the average of the variables at sigma levels from the ground to the mountain top at upstream points UP1 (MET1 case) and 2 (MET2 case). For the lowest mountain, F r ≈2 during nighttime and even higher at some other times during the day, indicating supercritical conditions. The Brunt–Väisälä frequency around mountain top is between 0.020 and 0.025s−1. The generated mountain waves have relatively short wavelength and small amplitude and a modest HJ that propagates downstream can be seen at the initial stages of the episode at 15:00 UTC 23 December (Fig. 5a). Some 12h later, during nighttime, the aforementioned strong jet extending for tens of kilometers downwind is fully formed with values above 35 m s−1at about 750m above ground and strong turbulence at the surface and in the layers above the jet, where stability is much reduced. The temperature profile upwind (at 03:00UTC 24 December and location UP1 in Fig. 5) shows in the temperature profile a stable lower boundary layer capped by a very stable isothermal layer from 850 hPa up to about 800hPa, or 2500m, defining the depth of the cold air pool. At the observation location MET1, downwind from the mountain, the lowest layers up to about 900 m are well mixed (slightly lower, up to 750 m further downstream), a result of the intense surface turbulence. Above the latter Earth Syst. Dynam., 10, 485–499, 2019 www.earth-syst-dynam.net/10/485/2019/ M. A. Prósper et al.: Downslope windstorms in the Isthmus of Tehuantepec during Tehuantepecer events 493 Figure 6. (a) Froude number upstream of the mountain, north of MET1 (UP1) and of MET2 (UP2). The gray zone represents nighttime. (b) Skew-T log-P diagram at MET1 and MET2 and (c) at their corresponding upstream points UP1 and UP2. Wind barb profile for each pair of points at right. height and up to about 1500 m elevation, a strongly stable layer exists corresponding with the aforementioned packing of the isentropes (and streamlines). This is where the highest wind speeds are found. Stability is sharply reduced aloft, in the region encompassed by the dividing streamline, and winds are rather weak and have an easterly component for the most part, parallel to the mountains. For the higher mountain north of MET2, F r ≈1 consistently in the period, indicating a critical flow regime prone to the formation of HJs (Vosper2006, 5B). The Brunt– Väisälä frequency around mountain top is between 0.012 and 0.015 s−1and the generated waves have higher amplitudes than in the MET1 case. Wave overturning and breaking is also much more pronounced (Fig. 5d, highlighted in white) and the well-mixed region forming to the lee of the crest is deeper. Isentropes and streamlines that sink underneath this region are packed in a very shallow layer on the lee slope of the mountain, generating an intense DSWS with speeds above 35 m s−1at the surface. These strong winds end abruptly at the foot of the hill, where the flow transitions to subcritical conditions and a marked stationary HJ forms, with vertical wind speeds of 6m s−1. A rotor extending from the jump to the location of observation station MET2 is also evident in Fig. 5l. The temperature profile upstream is very similar to that of the MET1 case, but downwind from the mountain at location MET2 lacks the strongly stratified layer www.earth-syst-dynam.net/10/485/2019/ Earth Syst. Dynam., 10, 485–499, 2019