Atmospheric Blocking : The Impact of Topography in an Idealized 1 General Circulation Model 2

14 Atmospheric blocking can have important impacts on weather hazards, but the fundamental dynamics of blocking are 15 not yet fully understood. As such, this work investigates the influence of topography on atmospheric blocking in terms of 16 dynamics, spatial frequency, duration and displacement. Using an idealized GCM, an aquaplanet integration, and integrations 17 with topography are analyzed. Block-centered composites show midlatitude aquaplanet blocks exhibit similar wave activity 18 flux behavior to those observed in reality, whereas high-latitude blocks do not. The addition of topography significantly 19 increases blocking and determines distinct regions where blocks are most likely to occur. These regions are found near high20 pressure anomalies in the stationary waves and near storm track exit regions. Focusing on block duration, blocks originating 21 near topography are found to last longer than those that are formed without or far from topography but have qualitatively 22 similar evolutions in terms of nearby geopotential height anomalies and wave activity fluxes in composites. Integrations with 23 two mountains have greater amounts of blocking compared to the single mountain case, however, the longitudinal spacing 24 between the mountains is important for how much blocking occurs. Comparison between integrations with longitudinally long 25 and short ocean basins show that more blocking occurs when storm track exits spatially overlap with high-pressure maxima in 26 stationary waves. These results have real-world implications, as they help explain the differences in blocking between the 27 Northern and Southern Hemisphere, and the differences between the Pacific and Atlantic regions in the Northern Hemisphere. 28 29 https://doi.org/10.5194/wcd-2020-2 Preprint. Discussion started: 16 January 2020 c © Author(s) 2020. CC BY 4.0 License.


Eulerian storm track 162
The storm track is presented as the standard deviation of a 24-hour difference of the daily mean Z500 field during winter 163 (Wallace et all. 1988). Consider 500 ( ) to be the daily mean Z500 value for an arbitrary gridpoint. To obtain the storm track: 164 1. The 24-hour difference, 500 , at each gridpoint is taken as: 165 500 = 500 ( + 1) − 500 ( ) 166 2. Then, the standard deviation of 500 for all winter timesteps at each gridpoint is taken to obtain the winter Eulerian 167 storm track value at that point. area grids before compositing. The initial time step of a block is the first timestep that the block satisfies the amplitude, size 192 and reversal conditions. The strongest time step of a block is defined as the time step with the greatest Z500' (at a single lat/lon 193 location) within a block. The final timestep is the last timestep a block satisfies the amplitude, size and reversal conditions. 194 The composites presented in this paper for the aquaplanet, unless otherwise stated, only include midlatitude-blocks 195 whose centroid are always south of 65˚ N. This is because we find that the high-latitude blocks exhibit distinct physical 196 behavior. The aquaplanet showed a greater tendency to produce more poleward blocks compared to the other configurations. 197 From reanalysis data, high-latitude blocks in the Southern Hemisphere have different dynamical evolution and different 198 impacts on the surrounding flow, as compared to midlatitude blocks (Berrisford et al., 2007). Based on these previous results 199 and our own findings, we present separate results for the midlatitude (count = 95; see Sect. 3.1) and high-latitude blocks from 200 the aquaplanet configuration (count = 46; see Sect. 3.2). The 65˚ N cutoff was chosen after estimates showed this to be near 201 the minimum in the meridional potential vorticity gradient, and thus the northern limit of the midlatitude waveguide (e.g. Wirth 202 et al. 2018). After changing the cutoff by +/-5˚, 65˚ N proved to be the best compromise between distinguishing dynamical 203 features between mid and high-latitude blocking, but also retaining enough members of each midlatitude and high-latitude 204 subset. 205 206

Separating blocks by region 207
To compare the dynamical evolution of blocks originating near the eastern edge of the ocean basins (denoted as East, 208 near the windward side of mountains and the high-pressure maxima of stationary waves) against blocks originating near middle 209 of the ocean (denoted as Mid, near the end of the storm tracks), blocks are sorted by their centroid location during their first 210 timestep. Each region spans 30˚-75˚ N for 100 degrees of longitude. In SymMtn, we defined our East region relative to the 211 mountain at 90E, which behaved similarly in our analyses to a region defined by the 90 W mountain instead. For the AsymMtn 212 configuration, East and Mid refer to two regions within the zonally larger ocean basin (which we refer to as the wide basin), 213 whereas blocks originating within the other ocean basin are only denoted by short basin. These regions are summarized in 214 2. Allowing replacement, randomly select a set of block durations within a given subset. The size of the random set 222 is given by the number of blocks in the subset being analyzed. 223 3. Place the durations yielded by step 2 into n equal sized bins (n=8 for figures in this paper) ranging from the 224 minimum to maximum duration of winter blocks between all model configurations. 225 4. Steps 2 and 3 are then repeated m times (m=1000 for figures in this paper) to produce an ensemble of m 226 probability density distributions for each subset. 227 5. For a given subset, the mean probability density distribution is computed by taking the mean of that subset's 228 distributions. This is then smoothed using a running mean. 229 6. For a given subset, the standard deviation of probability density distribution is computed by taking the standard 230 deviation of that subset's distributions 231 The results of this paper are nearly constant with respect to changes in the values of n (+/-2) and m (+/-200). For all 232 configurations, distributions and mean values presented for duration exclude any high-latitude blocking (blocks whose centroid 233 are ever poleward of 65° N). 65° N was found to be the most appropriate cut-off in each configuration for the same reasons as 234 described for the aquaplanet compositing. 235 236

Block displacement 237
To measure the propensity for individual blocks to move horizontally, we define a block displacement metric. In this 238 metric for an arbitrary block, the great circle distance between the block centroid at successive timesteps is computed. The 239 block displacement for each block is the sum of all displacements computed throughout its lifecycle, divided by the number 240 of timesteps (i.e. the average centroid displacement every 6 hours). 241 242

Statistical significance 243
To compare block frequency between configurations, the area-weighted mean of winter block frequency is computed 244 for each year of a given configuration. A 2-sample t-test is used on the yearly area-weighted mean values between 245 configurations to test for significant differences. Between configurations and regions, significance testing discerning mean 246 block duration and displacement employ a 2-sample t-test. An 85% confidence interval is imposed as the significance threshold 247 for all significance testing. 248 249 3 Results 250

Blocking in the aquaplanet 251
According to our tracking algorithm, there are blocking events in the aquaplanet integration, which agrees with 252 previous idealized modeling work (Hu et al., 2008;Hassanzadeh et al., 2014). An example of the beginning of a blocking 253 episode in the aquaplanet can be seen in Figure 2. Upstream and coincident with the block, a Rossby wave pattern can be observed in both the Z500 and Z500' fields ( Fig. 2 -the Z500 contours show a wave-like feature, and the Z500' field shows 255 an alternating pattern of low and high anomalies in the zonal direction). The presence of these features during the formation 256 of a block agrees with previous work for both simplified (Berggren et al., 1949;Rex, 1950;Colucci, 1985 We use block-centered compositing analysis to confirm that, on average, the blocks identified in the aquaplanet model 262 evolve in a dynamically similar manner to results shown in previous studies. Figure 3 shows block centered composites of 263 Z500' and → for the aquaplanet blocks in the midlatitudes (i.e., occurring between 30˚ and 65˚ of latitude). The onset of 264 blocking in the composite is similar to that found in the case study (Fig. 2): a Rossby wave train with low-pressure centers 265 upstream and downstream of the composite block centroid, and a large concentration of → upstream and entering the block 266 ( Fig. 3a). For composites over blocks at maximum strength, the wave pattern is no longer pronounced and low pressure is 267 concentrated equatorward and downstream of the block (Fig. 3b). Large magnitude → are concentrated inside the block during 268 this time (Fig. 3b). On the final day, the composite block's Z500 anomaly weakens, and low-pressure is concentrated 269 downstream from the block (Fig. 3c). Weak values of (2) The geopotential height field shows the evolution of a wave train that eventually dissipates as → are passed downstream 276 ( Fig. 3a-c). On the other hand, the high-latitude blocks from the aquaplanet display quite different behavior. 277 278

High-latitude blocking 279
As discussed in the methods section, the aquaplanet configuration has a larger amount of high-latitude blocking than 280 the other configurations (Fig. 4a, Table 2). These blocks have multiple unique characteristics, as compared to blocks from all 281 model configurations (including the midlatitude blocks for aquaplanet).  Overall, case studies and block-centered composites show that blocks in the idealized model share similar 296 characteristics and dynamics as blocks observed in reality. This holds even when blocks are sorted between midlatitude and 297 high-latitude events. Confident with the representation of blocking in the idealized model, next we focus on the climatological 298 response of blocking to topography. 299 300

The effects of topography on winter blocking 301
This section focuses on the effect of topography on climatological flow features and blocking climatologies. As 302 motivation, we first present results from reanalysis that agree with previously published studies. Then, we investigate the 303 response of the same climatological features in the idealized model to changes in topography. 304 305

Reanalysis 306
The different topographic configurations of the northern and southern hemispheres produce distinct spatial 307 distributions of general circulation features and atmospheric blocking. Figure 5 shows the stationary wave, U250 climatology, 308 storm track and blocking climatology for winter in ERA-Interim. Stationary wave patterns can emerge due to land-sea heating 309 contrasts and flow deflection by orographic geometry. The two strongest regions of anomalous high-pressure in the Northern 310 Hemisphere (NH) are located on the windward side of the Rocky Mountains, and near the western edge of Europe (Fig. 5a). 311 The high near the Rockies is part of a wave train induced by the mountains (e.g., White et al., 2017). The high near Europe is 312 more likely driven by land-sea contrast. The Asian orography also produces a stationary wave response and is an important Rocky Mountains (Fig. 5a and Fig. 5c). This region is also spatially overlapping with the Pacific storm track exit. For the NH 325 Atlantic basin, the location of the blocking maximum and high-pressure stationary maximum are within close proximity, but 326 both the storm track exit and maximum spatially overlap with them ( Fig. 5a and Fig. 5c). In the NH, blocks rarely occur near 327 the low-pressure anomalies of the stationary wave ( Fig. 5a and Fig. 5c). 328 In the Southern Hemisphere (SH), the high-pressure maximum is more poleward than the Northern Hemisphere 329 maxima and stretches from the southwestern tip of South America into a secondary maximum southeast of Australia (Fig 5b).

330
This matches what is reported in Quintanar and Mechoso (1995). Stationary wave features are far less apparent in the Southern 331 Hemisphere, presumably because of the relative lack of topographic forcing compared to the Northern Hemisphere. 332 The lack of topographic forcing in the SH allows there to be one distinct band of maximum U250 (Fig 5b). The U250 333 maximum in the SH stretches from the Indian Ocean into the Pacific and maximizes East of Australia (Fig 5b). The single 334 storm track maximizes in the Indian Ocean near Antarctica and stretches from the Atlantic to the Pacific (Fig 5d), far upstream 335 from the region of maximum U250 (Fig 5b and Fig 5d). The SH storm track is as reported in Nakamura and Shimpo (2004). 336 In the Southern Hemisphere, our blocking climatology is similar to that reported in Brunner and Steiner (2017). The 337 blocking maximum is near the high-pressure anomaly of the stationary wave and the exit region of the Pacific storm track of 338 the Southern Ocean ( Fig. 5b and Fig. 5d). The spatial frequency of blocking in the SH extends into the SH Atlantic storm track 339 entrance region, away from the high-pressure anomaly, but the local blocking maxima in the SH Atlantic is weak compared to 340 the SH Pacific maxima (Fig. 5d). 341 Topographic differences yield contrasting spatial distributions of stationary waves, U250, storm tracks, and blocking 342 between the hemispheres. These observations lead to the specific questions this subsection seeks to address: 343 -What effect does topography have on blocking? 344 -What role do stationary-waves and storm track exit regions have in setting the locations and intensity of blocking 345 maxima? 346

Blocking in idealized model experiments 347
The idealized model configurations allow us to systematically investigate the response of atmospheric circulation and 348 blocking to topography. As we did for reanalysis, for each model configuration we examine the stationary wave, U250, storm 349 tracks, and the blocking climatology. 350 As expected, a stationary wave is absent in the aquaplanet (Fig. 6a), and upon introducing topography, zonally 353 asymmetric forcing is imposed, and a stationary wave is induced (Figs. 6b-6d). SingleMtn contains a high-pressure anomaly 354 near the coastline on the windward side of the mountain, and a low-pressure anomaly on the leeward side (Fig. 6b). This results 355 in a meridionally tilted stationary wave pattern that extends into the subtropics leeward of the mountain. This pattern has been 356 explained in previous idealized modeling work (Grose and Hoskins, 1979;Cook and Held, 1992;Lutsko 2016). The high-357 pressure anomaly extends approximately 180° of longitude upstream of the mountain and weakens from east to west. 358 In SymMtn, the configuration with two mountains and equal-sized ocean basins, each mountain induces a 359 meridionally tilted stationary wave pattern (Fig. 6c) similar to that in SingleMtn (Fig. 6b). The zonal extent of the high-and For AsymMtn, the placement of the topography creates two ocean basins of different zonal extents: a short basin and 364 a wide basin. Like SymMtn, each mountain in AsymMtn induces a meridionally tilted stationary wave, however, the 365 asymmetric configuration of the mountains results in asymmetric zonal extent in the anomalies (Fig. 6d). In the short basin, 366 the anomalies have less zonal extent than those in SymMtn, and the opposite holds true for the wider basin. Further comparing 367 the two basins, we find the high-pressure anomaly in the wide basin extends 100 degrees westward from the mountain, much 368 farther than that of the short basin. This extended high-pressure anomaly is related to blocking, which we will further address 369 below, but first we analyze U250. 370 371 250 hPa zonal wind climatology 372 In the aquaplanet, U250 is zonally symmetric. When topography is added, localized regions of U250 maxima occur. 373 In SingleMtn, the U250 maximum occurs on the leeward side of the mountain, equatorward of the low-pressure anomaly (Fig.  374 6b). The stationary wave pattern associated with the topography generates cold advection towards the southeast on the lee of 375 the mountain. This is due to both the change in wind direction created by the mountain and the differences in heat capacity for 376 the topography as compared to the ocean. The cold advection leads to a local maximum in meridional temperature gradient 377 east of the topographical feature (not shown). Related to this temperature gradient, the U250 maximum must exist due to 378 thermal wind balance. This pattern of the U250 maximum occurring just downstream of mountains is the same as what occurs 379 for the NH in observations (Fig. 5a). 380 In SingleMtn there is also a relative suppression of U250 nearly 120° downstream of the mountain, from about 150° 381 W -110° W, followed by a secondary maximum of U250 from roughly 110° W -0°. The disjointed distribution of U250 is 382 perhaps a consequence of blocking and will be discussed further in the blocking climatology subsection. 383 The U250 maxima for SymMtn and AsymMtn are located on the poleward side of the low-pressure anomalies of each mountain is suppressedlikely because of influence from the downstream mountain, and consistent with the zonal suppression 387 of the stationary wave (Fig. 6d). 388 The U250 field acts as a waveguide for synoptic scale Rossby Waves -e.g. The storm track in the aquaplanet is zonally symmetric (Fig. 6e), while the topographical configurations have zonally 394 asymmetric storm tracks whose locations are set by mountains (Figs 6f-6h). In the topographic configurations, the storm tracks 395 almost exactly overlap with the U250 maxima, with the exception that the storm track maxima are slightly upstream from the 396 U250 maxima (Figs 6b-6d, Figs 6f-6h). In AsymMtn Short Basin, the zonal extent of the storm track is suppressed by the 397 topographical spacing, similar to U250 and the stationary wave. The blocking climatology in the aquaplanet is not zonally symmetric for the 30-year integration (years 11-40; Fig.  405 6e). For a 300-year integration, the climatology is much closer to being zonally symmetric, though it has still not converged 406 (not shown). No zonal asymmetries in forcing exist in the aquaplanet, so the zonal asymmetries attest to the internal variability 407 and relative rarity of blocking events identified in the aquaplanet. The configurations with topography are closer to reaching 408 convergence after 30 yearsin terms of the local maxima occurring just west of the topography. To demonstrate this, we 409 compare climatologies of Aquaplanet and AsymMtn based on randomly chosen subsets of years. Aquaplanet blocking 410 climatologies using randomly sampled years produce results with varying spatial frequencies (Fig. 7). 411 Upon adding topography, spatial maxima form in the blocking climatology (Figs. 6e-6h) and significantly more 412 blocking occurs overall (see Table 2 for quantitative differences). The result that adding topography to our model leads to 413 greater block frequency matches with observations, since the Northern Hemisphere contains a relative abundance of 414 topography and blocking, when compared to the Southern Hemisphere (Figs. 5c-5d). 415 The blocking maximum in SingleMtn (Fig. 6f) is slightly upstream from the maximum high-pressure anomaly (Fig.  416 6b), on the windward side of the topography. This is similar to the NH Pacific blocking maximum being situated northwest of 417 the Rocky Mountains in observations (Fig. 5c). The high-pressure anomaly on the windward sides of mountains acts as a 418 source region of anticyclonic vorticity and can be recognized as ridges in instantaneous maps of geopotential height. These ridges serve as precursors for topographically induced blocks, which are then amplified and maintained by transient eddies 420 and → (Nakamura et al., 1997; TN01). 421 A secondary blocking maximum in SingleMtn is found towards the western end of the high-pressure anomaly, near 422 the storm track exit (Fig. 6f). A tertiary, but relatively weak blocking maximum is found from roughly 150° W -110° W, 423 where U250 contains a local minimum in between the two U250 maxima. The blocking in this region is a probable explanation 424 for the gap in the U250 maximum, as blocks are known to inhibit or even halt zonal flow. The second and third blocking The presence of a second, symmetrically placed mountain in SymMtn leads to the occurrence of significantly more 429 blocking than in the aquaplanet, SingleMtn, and even AsymMtn ( Fig. 6g and Table 2). The blocking maxima in SymMtn sit 430 near the intersection of the high-pressure anomaly and storm track exit (Fig 6c and Fig. 6g). In AsymMtn there are blocking 431 maxima also on the windward sides of the mountains near each respective high-pressure anomaly (Fig. 6h), and the overall 432 area-averaged block frequency is slightly greater than SingleMtn, but less than SymMtn (Table 2). 433 In AsymMtn Wide Basin, the blocking maximum is close to the stationary wave maximum and a secondary blocking 434 maximum occurs at the western edge of the high-pressure anomaly, near the storm track exit (Fig. 6d and 6h). As in SingleMtn, 435 these separate maxima correspond to distinct block genesis regions. 436 In AsymMtn, the short basin has a greater block frequency maximum than wide basin (Fig. 6h). Like SymMtn, the 437 short basin in AsymMtn has a storm track exit region that overlaps with the high-pressure maximum of the stationary wave 438 when compared to SingleMtn and AsymMtn Wide Basin. This perhaps explains the enhanced blocking climatological 439 maximum in AsymMtn Short Basin compared to AsymMtn Wide Basin. On the other hand, AsymMtn Short Basin has such a 440 small zonal extent that the storm track exit overlaps with the mountain. Thus, in this short basin there are more likely to be 441 times in which storm development occurs just upstream of the mountainand such conditions would inhibit blocking. 442 As observed in the Atlantic basin of Earth, we suspect the shortened jet in AsymMtn Short Basin acts as a waveguide 443 that funnels transient eddies and → into the anticyclonic anomaly of the stationary wave, and these eddies have the potential 444 to feed blocks or help destroy them. The details of those processes are a focus of future work. 445 When comparing the blocking climatologies for each configuration, we find that blocks are predominantly generated 446 at high-pressure stationary maxima, regions dominated by wave breaking (storm track exit), or at some spatial mixture of the 447 two (Figs. 6e-6h). The aquaplanet shows that blocks can arise purely from eddy-eddy interactions, whereas the other 448 configurations show that blocks can also be induced by topography, at a more frequent rate. 449 We want to highlight the result that SymMtn has the largest area-averaged block frequency (Table 2) and number of 450 events (Table 3) out of all the configurations. We hypothesize that this is because the ocean basins in SymMtn have a zonal 451 extent that allows a synergy between the block genesis mechanisms associated with the high-pressure anomaly induced by the topography and block maintenance mechanisms associated with the storm track exit. A similar inference can be made when 453 comparing the short and wide basins of AsymMtn, where the short basin contains a stronger blocking maximum and a more 454 spatially coincident storm track exit with the high-pressure of the stationary wave. However, as mentioned above, AsymMtn 455 Short Basin is so short that there is not enough spatial separation between the storm track entrance and the downstream 456 topographical feature. The processes governing the interactions of the storm tracks and the topographical features in relation 457 to blocking are topics of future work. 458 459

Block duration and displacement 460
One of the characteristics that allows blocks to influence midlatitude weather is their persistence. As such, we examine 461 the influence of topography on block persistence using our duration metric. First, we find that adding topography, regardless 462 of configuration, leads to longer duration blocks on average (Table 3). For SingleMtn, the difference compared to Aquaplanet 463 is statistically significant (8.4 versus 7.3 days, respectively). For SymMtn and AsymMtn, the mean duration is longer, but the 464 difference is not significant at the 85th percentile. This is because of the large variance in block duration when we consider all 465 midlatitude blocks generated by the model. However, if we subset the blocks, based on the location in which they are generated, 466 this result changes. 467 Using the regions defined in Table 1, we found that for blocks that occur in the eastern portion of the ocean basins 468 (i.e., East Blocks), the mean durations are all significantly longer than those of aquaplanet (Table 4, and see Fig. 8.b. for the 469 probability density distributions). The east portion of the ocean basins is near the local maxima in the stationary wavewest 470 of the topography, thus, the longest duration blocks per configuration are those that are generated just upstream of topography. 471 Furthermore, the average duration values for the East blocks in SingleMtn and AsymMtn are greater than their Mid 472 counterparts (i.e. blocks that start near the storm track exits, see Table 1 and Fig. 6). SingleMtn East and SingleMtn Mid have 473 mean block durations of 9.1 and 8.2 days respectively (Table 4). AsymMtn Wide Basin East and AsymMtn Wide Basin Mid 474 have significantly different mean block durations of 8.3 and 7.1 days respectively (Table 4). Thus, blocks that form near 475 topography in this model (i.e., the East blocks), tend to persist for longer times than those that form far from topography (i.e., 476 Mid blocks, or blocks in the aquaplanet configuration). The same analysis applied to the NH and SH in reanalysis found that 477 the NH, which presumably contains much more topographically forced blocks, has an average block duration (8.0 days) that 478 is significantly longer than those from the SH (6.9 days). 479 A natural question to ask then is: Why do blocks originating near topography have longer durations than those in the 480 aquaplanet? Given that the topographic configurations contain both stronger localized storm track regions and blocks generated 481 by topography, whereas the aquaplanet does not, we hypothesize two possible explanations for the differences in duration: 482 1. The stronger localized storm tracks create more eddies, which would provide more transient eddies that could 483 feed the blocks through dry dynamics (e.g., Shutts, 1983; TN01; Yamazaki and Itoh, 2013) or moist dynamics 484 (Pfahl et al., 2015).
2. Topographically generated blocks last longer than blocks predominantly generated by eddy-eddy interactions 486 because they are fundamentally different. 487 Regarding Hypothesis 2, we analyzed → composites for East Blocks as compared to Mid Blocks and found minimal 488 differences aside from increased composite → magnitudes for the Mid Blocks (Fig. 9). Since the Mid blocks are those more 489 likely to be generated by eddy-eddy interactions, this result refutes Hypothesis 2. Related to this, as discussed in Sect. 3.1, the 490 life cycle composites of the wave activity flux are very similar for the aquaplanet configuration and the configurations with 491 topography (i.e . Fig 3 and Fig. 9). These results point more toward hypothesis 1 but are very much preliminary. More work is 492 planned to investigate the maintenance of blocking between the model integrations. 493 Next, we test for differences in block displacement to determine if topography obstructs the movement of blocking 494 events. The differences in average block displacement (Table 3)

Summary and conclusions 500
This work utilizes an idealized moist GCM to better understand atmospheric blocking. We start with an analysis of 501 blocking in an aquaplanet. Then we systematically add topographic features to investigate the influence of topography on 502 blocking, in terms of their climatological location, duration, and displacement. 503 In the aquaplanet we find that blocks can be generated purely through eddy-eddy interactions; i.e., they do not require Like Berrisford et al. (2007), who looked at blocking in Earth's Southern Hemisphere, we identify distinct high-511 latitude blocking events that differ from midlatitude blocks. High-latitude blocks in the aquaplanet have lower geopotential 512 height anomalies, primarily occur poleward of the main zonal channels of → , and do not contain strong concentrations of → 513 at peak strength. This suggests an alternative maintenance mechanism for high-latitude blocks than those proposed for blocks 514 in general by Nakamura et al. (1997) and TN01. High-latitude blocks are also identified in the topographic configurations, but 515 to a lesser extent than the aquaplanet. 516 https://doi.org/10.5194/wcd-2020-2 Preprint. Discussion started: 16 January 2020 c Author(s) 2020. CC BY 4.0 License.
For the topography experiments, we modified the aquaplanet model in the following ways: (1) adding a single 3-km 517 mountain; (2) adding two 3-km mountains evenly spaced with respect to longitude; and, (3) adding two 3-km mountains 518 asymmetrically space with respect to longitude, to create one long and one short ocean basin. 519 The addition of topography led to some changes in blocking that were universal across all configurations with 520 topography, compared to the aquaplanet integration: 521 -There are localized maxima in blocking, upstream of topography; near the high-pressure maximum of the stationary 522 waves; a source region of anticyclonic vorticity. 523 -There is significantly more wintertime blocking overall with topography present. 524 -When topography is present, blocks have longer durations. 525 -Topography does not play a key role in determining the characteristics of block movement. 526 Based on ERA-Interim reanalysis, these results mirror what is observed for the NH and SH, where the NH contains more 527 topography, blocking, and longer lasting blocks. 528 The addition of topography also induces stationary waves, and localized maxima in the jet streams and the storm 529 tracks. This response has been documented previously, but our interest was the interaction between these features and blocking. 530 In all configurations with topography, local blocking maxima are found near high-pressure stationary anomalies as well as 531 storm-track exit regions, where Rossby waves tend to break. A local minimum in blocking is coincident with the jet stream 532 maxima and storm track entrance regions. 533 The spacing between the two mountains is important for the amount of blocking that is produced: symmetrically 534 placed mountains leads to significantly more blocking than all other configurations. Both blocking maxima in SymMtn 535 spatially overlap with their ocean basin's respective storm track exit region and anticyclonic stationary anomaly. We suspect 536 SymMtn's increased block frequency reflects a spatial resonance between breaking Rossby waves at the storm track exits 537 interacting with high-pressure anomalies generated by the mountains. This helps explain some of the differences in the 538 blocking climatology we observe between the Pacific and Atlantic in the NH. 539 Though the blocking maxima in the NH Atlantic and Pacific basins are similar in magnitude, the Pacific maximum 540 covers a larger areathus there is more blocking in the Pacific. In the NH Pacific, a similar spatial distribution to SymMtn is 541 observed between the storm track exit, blocking maximum and stationary wave induced by orography. The Atlantic on the 542 other hand, is akin to the Short Basin in AsymMtn, a storm track whose exit and maximum both are semi-coincident with the 543 stationary high-pressure and blocking maximum. Our results suggest the broader Pacific blocking maximum is a consequence 544 of better spatial resonance between the Pacific storm-track exit and stationary anomaly, compared to the Atlantic. An 545 alternative hypothesis is that the semi-coincidence between the storm track and blocking maxima in the Atlantic inhibits 546 blocking. Another possible explanation is that the stationary wave in the Atlantic is forced by land/sea contrasts rather than a 547 mountain, leading to different interactions with its storm track, as compared to the Pacific. Further work will be done to 548 investigate the sensitivity of climatological blocking maxima to the location of storm track exits.
In the configurations with topography, blocks generated near topography last longer, on average, than those produced 550 away from topography. However, compositing results of Z500' and → found blocks forming near and away from topography 551 yielded little differences aside from blocks away from topography interacting with larger magnitudes of → compared to near 552 topography counterparts. Further work is planned to provide a mechanistic explanation for these differences we find in block 553 duration. 554 Overall, this work elucidates fundamental information on the formation, dynamical evolution, spatial distribution, 555 duration, and displacement of atmospheric blocking. Future work will utilize a suite of dynamical diagnostics to take a deeper 556 look into the differences between blocks generated near topography compared to those that are not, and how it relates to what 557 is observed in reality. 558 559 https://doi.org/10.5194/wcd-2020-2 Preprint.