Annular Solar Eclipse – October 14, 2023

The Eclipse Weather Desk

The Eclipse Weather Desk provides quick links to satellite imagery and numerical weather models ahead of eclipse day to aid in your planning and final site selection. Click on the title or here to go to the site.

The annular eclipse of 14 October, 2023, is the second of three eclipses that cross the United States in the eight years between 2017 and 2024. In contrast to the 2017 total, this eclipse is shared with portions of Central and South America. The shadow path’s large range of latitude, from 50 degrees north to just below the equator, dictates a spectrum of climates and weather conditions. Topography—particularly the Cascade Range in the American Northwest, the Tabasará Mountains in Panama, and the Andes in Colombia—plays an important roles in moderating the cloud cover as the air rises and falls in its passage over the terrain (Figure 1).

October 14 morning brings the shadow’s first contact with the Earth, in the Pacific west of Vancouver Island and north of the semi-permanent Pacific high-pressure anticyclone that lies off California (Figure 2). A few minutes later, it reaches the coast of Oregon, crossing a narrow sandy beach south of the community of Florence and heading inland to ascend the 1000-m heights of the Coast Range and shortly afterward, the 2400-m peaks of the Cascade Range. From there, the up-and-down high plains of Nevada’s Great Basin lead the shadow past Four Corners to Texas and the low-lying Gulf Plains before it exits onto the Gulf of Mexico. October is a friendly month and much of the American part of the track offers very promising prospects of sunny skies on eclipse day.

Figure 2: Average October sea level pressure along the track of the eclipse.

Across the western Gulf of Mexico, cloudiness is inhibited by the sinking air of the sub-tropical Atlantic anticyclone, which descends from higher layers to dry the atmosphere and banish clouds. There are some flies in this ointment however—the threat of a hurricane is one of them and passing cold fronts another—but for the most part, the warm waters are dotted by small convective clouds that pose little threat to an eclipse. It is only when the track reaches the Yucatan Peninsula and beyond that cloudy skies become a significant threat.

The high Sun of the tropics is a potent force in warming the ground and initiating the growth of convective clouds. Over the Yucatan, where the influence of the Atlantic and Pacific anticyclones still have a little currency, the convective clouds are primarily small and easily dissipated. Further south, however, the track begins to encounter the higher humidities and mountains of Central America and in Panama, the Intertropical Convergence Zone (ITCZ).

The ITCZ is the Earth’s “weather equator” where winds from the Northern and Southern Hemispheres collide and rise upward, forming a persistent band of cloudiness and precipitation that can be traced around the globe. Over the oceans, the ITCZ forms an easily identified band of cloudiness but over Panama and Colombia’s mountains and Brazil’s Amazon Basin, it broadens into a continent-sized low-pressure regime that blossoms with dense convective clouds nearly every day.

As the day moves toward late afternoon, the lunar shadow moves into Brazil and samples the equatorial moisture of the Amazon Basin. October is the first month of the wet season over the Amazon and the cloud cover is heavy and persistent. Toward the end of the track, with the Sun less than 10° above the horizon, the eclipse track moves onto the higher plateau of Northeastern Brazil where the seasons are reversed: October is the driest month of the year, and much of the cloudiness melts away, leaving favourable weather prospects as the eclipse sinks onto the western horizon.

The United States

The eclipse path through the United States begins and ends in modestly cloudy climates, but the midriff welcomes skies with generous sunshine (Figure 3)

graph of centreline cloudiness
Figure 3: Graph of mean October cloud cover (2000-2018) acquired during the satellite pas at 10:30 am along the centreline of the eclipse track. Units are fractional sky cover, which can be interpreted as percent cloud cover.

Oregon

The lunar shadow comes arrives at the Oregon beaches in familiar country, atop many of the communities that greeted the arrival of the 2017 total eclipse. The season is two months more advanced this time, and monthly precipitation has passed it summer minimum and begun to climb to its winter peak in December and January. The cloud and precipitation patterns in Oregon are more complex than in other states along the shadow path because of the influence of the Coast and Cascade mountain ranges (Figure 4), which restrict the influx of moisture from the Pacific Ocean. The ups and downs of these two ranges creates several microclimates that may be used to benefit the eclipse watcher.

map of Oregon
Figure 4: Topographic map of Oregon and Northwest Nevada along the eclipse track.

One of these cloud refuges is the Willamette Valley, which stretches southward from Portland to Eugene before tapering off into a series of narrow river valleys that extend to Grant’s Pass and beyond. With the Coast Range on one side and the Cascade Range on the other, the valley has a well-deserved reputation for a dry and sunny environment. Unfortunately, the tucked-in valley bottom is plagued by overnight and morning fog in October, patches of which often linger past the noon hour—well past eclipse time. Fog is most frequent on the north side of the shadow track, so if you aren’t prepared to wait and hope that it burns off in time, a quick run southward down Interstate 5 may break out into clearer skis.

Figure 5: Average October cloud fraction  at 10:30 am for Oregon and Nevada based on 19 years of observations from the Terra satellite, 2000-2018.

The advantage gained in the Willamette Valley is small, about a 4 percent reduction in average cloud cover according to Figure 3’s graph. The cloud map in Figure 5 reveals that the cloud-eating effect of the Coast Range is greatest to the south of the central line and fades away to the north.

Past the Willamette Valley, cloud cover begins to decline, though there is a small upward bump over the Cascades before the track moves onto the Columbia Plateau and the High Desert. Past

Crater Lake, average monthly cloudiness drops below 50 percent and continues to fall, irregularly, to the Nevada border and beyond. Small rises in the terrain are matched by small upward bumps in cloudiness, but the High Desert offers the best eclipse-watching prospects in Oregon. Within the 40-to-45 percent average cloudiness in the area, the countryside near Klamath Falls and around Denio offers the finest cloud prospects for long-range planning.

Climate data for places in Oregon
Table 1: Climate statistics for selected sites in Oregon. % Sun refers to the measured average October percent of maximum daily sunshine (sunrise to sunset) .

Eclipse-viewing expeditions in 2023 would be best served by locating east of the Cascades on the south side of the centre line. The terrain in this area is open and sparsely treed, giving an excellent view toward the Sun at the moment of central eclipse. Photographers who wish to gamble with the weather gods may wish to settle alongside 1800-m-high Crater Lake, lying just south of the centre line in the Cascades. The crater and its lake will frame a 19-degree-high Sun. Be warned, however, that winter starts early at high altitudes, and Crater Lake records an average of 1.9 inches (46 mm) of snow in October (Table 1).

Lincoln Beach, Oregon, inside the north limit where the eclipse comes ashore. Lincoln Beach was also beneath the shadow at the eclipse of 2017.

For those who want to be first in line for the shadow, the coast of Oregon offers a visitor-friendly environment with plenty of accommodation—and familiarity with eclipses from the 2017 experience. There is not a lot of room to move if the weather is unfriendly however, as highways inland are limited and the slopes of the Coast Range are cloaked in dense forests with very little view to the rising Sun. Even in the Willamette Valley, vistas must be chosen carefully, particularly south of Eugene, where the terrain closes in. The same limitations apply to the slopes of the Cascades, which do not begin to open up until Crater Lake has been left behind and you’ve reached Highway 97. 97 provides a convenient path to run north and south across the eclipse track if weather becomes a problem on October 14.

A diversion into California

The small wedge of the eclipse path in California crosses a high semi-desert with a rolling topography. Cloud prospects are similar to those in nearby Oregon, with average amounts ranging from 38 to 46 percent; the lowest values are along the south limit. Here again, cloud cover mimics the terrain, with higher amounts along the Nevada border east of Alturas and Goose Lake where the elevation climbs above 2200 m in the Warner Mountains.

October marks the beginning of the winter wet season in Alturas but the precipitation for the month is a barely noticeable 10 mm that falls on an average of 5 days. Unfortunately, the eclipse duration is short, just under 2 minutes at Alturas, which lies only 10 km from the south limit. The area would be suitable for those wishing to be at the limit with its spectacular display of Baily’s beads.

Great Basin Country: Nevada and Utah

As the eclipse track passes the Warner Mountains and crosses into Nevada, it enters the Great Basin, a large elevated plateau that covers most of Nevada, half of Utah, and parts of the surrounding states. The Sierra Nevada and Cascade Mountains block the flow of Pacific moisture from the west and the Wasatch Mountains do the same to the east, casting a large rain shadow over the entire region (Figure 6). The Basin has a closed drainage system, holding all of its precipitation (which comes mostly as winter snow) internally in underground storage or in small lakes.

Map of Nebraska and Utah
Figure 6: Topographic map of Nevada and Utah along the eclipse track.
Figure 7: Average October cloud cover for 10:30 am local time derived from 19 years of observation from the Terra satellite . Data: NASA.

The Great Basin is a washboard of parallel mountain ridges and valley bottoms (Figure 6), a feature known as basin-and-range topography. A similar pattern is evident in the cloud map, though it takes some effort to fit the cloud pattern to the rise and fall of the terrain. Cloud cover along the eclipse track comes from the influence of several weather patterns:

  • Storm systems moving off of the Pacific that drag a cold front over the mountains and into Nevada
  • Storms moving onto the southern California coast that spread high-level cloud northward
  • Wave clouds that form in an upper-level westerly flow over the Sierra Nevada mountains in California and spread eastward
  • Convective clouds that form on the mountain ridges in the unstable air after a cold front passes

For the eclipse traveller, the Basin is a compelling destination. Average monthly cloud cover in October ranges from a high of around 45 percent along the Oregon-California border to less than 30 percent in southern Utah though there are many ups and downs along the way, as shown in both the centreline graph (Figure 3) and in the map of Figure 7. As in Oregon, the most reliable weather is found on the south side of the track.

Storms that move off of the Pacific generally head northward into Canada, but many of them drag a long tail of cold-front cloudiness that sweeps over the mountains into Nevada. The mountains wring a lot of moisture from the front, turning weaker systems into a band of semi-transparent mid- and high-level cloud. Stronger systems bring heavy overcast and precipitation. The cold front leaves an unstable atmosphere when it departs, causing an explosion of small to medium convective clouds on the following sunny day when the Sun heats the ground. Because the mountain ridges are darker than the valley bottoms, the ground heats most on the higher terrain, forming clouds there and leaving the lower elevations in sunshine. Later in the day, the ridgeline clouds may expand out across the whole of the landscape, but the eclipse will come and go before these clouds have had a chance to spread.

When winds flow over a mountain barrier at right angles, a series of orographic waves may form in the downstream atmosphere, forming bands of clouds that extend for a long distance to the east. The clouds develop in the crest of each wave and dissipate where the air descends into the troughs. The separation of the individual wave crests depends on the wind speed at mountaintop, but once established, tends to be relatively stable, with persistence in the same place over an hour or more—sometimes many hours. If orographic waves form on eclipse day, you can be reasonably assured that a location in the clear skies between the waves will persist for enough time to watch the event. The greatest difficulty might be in finding an opening big enough to be able to see a sun that is only 25 degrees high.

Over the Great Basin, it is the Sierra Nevada Mountains along the California border that are the biggest wave generators so Nevada is more affected than Utah because of its proximity to the Sierra. Waves come most often with southwest winds in the upper atmosphere (southwest winds are perpendicular to the mountains), so they are more frequent and usually thicker in the northern parts of the Basin. That southwest wind flow does not extend into southeast Utah, making wave clouds unlikely there and helping to give the area its very favourable cloud climatology. If wave clouds are expected on October 14, a move toward the south limit, especially in Utah, should bring you into better skies. Satellite images on the day will allow you to evaluate the situation.

In order not to put too much of a damper on the prospects for Nevada and Utah, it should be noted that of the 20 October 14s from 2000 to 2019, 16 would have provided an easy and successful view of the eclipse, though sometimes through thin cloud or requiring a small movement away from hills. The remaining days offered more limited prospects, but on only one day was Utah nearly completely clouded out.

Climate data or Nevada
Table 2: Climate data for selected sites in Nevada.

The Great Basin’s semi-desert character speaks of a relatively light annual precipitation, so even though October comes at the start of the winter wet season, rain and snowfall accumulations are relatively small. In Table 2, we see that average October morning temperatures are below the freezing point in Nevada and close to it in Southern Utah. Precipitation is relatively light for the month and the number of days with precipitation, including snow, is small. Average sunshine figures seem generous in light of the satellite-based cloud estimates, but much of the difference may be due to the high frequency of thin cloud formed in the flow over the mountains. Sunshine observations were largely terminated in the United States in the 1990s and some of these statistics are now more than 30 years old.

Last minute movements to seek better weather conditions in Nevada will have a few route choices. Interstate 80, from Reno to Salt Lake City, angles across the track for a distance of 450 km from Unionville to West Wendover and its 80 mph (128 km/h) speed limit allows for quick relocation. Since better conditions are, on average, to the south of the track, the Interstate does not allow quick repositioning along a north-south axis. A better solution for a cross-track change is Interstate 15 in Utah, which runs almost directly across the track, from Levan to Cedar City.

Another alternative is U.S. Route 50, through Eureka and Ely in Nevada to Delta and Levan in Utah. This route, known colloquially as “the Loneliest Road in America,” runs more-or-less parallel to the track in Nevada before turning northeastward in Utah. Because it stays on the south side of the eclipse track in Nevada, it has the advantage of sampling the part of the track with the greatest likelihood of good weather.

Bryce Canyon, Utah

A good “first base” might be Ely in Nevada, which lies at the junction of Route 50 and Route 93. The first allows for an east-west run, while the second heads north across the track. Ely is far enough from the Sierra Nevadas to avoid some of the wave clouds that might be generated in the flow over the mountains, it lies in the climatologically more favourable south side of the track, and movement to the centre line involves only a one hour (75 km) trip. In Utah, Richfield, near the junction of Interstate 70 and Interstate allows a quick cross-track adjustment but is limited in the along-track direction by the angle that Interstate 70 takes on its way to the north. In Utah, it’s probably best to head to the southeast corner of the state where Mother Nature promises the best cloud prospects.

The best advice that can be offered for the Great Basin states is to be prepared to go south—either toward the south limit or to Southeast Utah—if the weather forecast is unfavourable. The transition from the Great Plains climate to the Southwest Desert climate occurs on the east side of the Sevier and Wasatch Plateaus (from Bryce Canyon onward) and cloud amounts beyond that point are very good. For an extra treat, both states have natural attractions in world-famous National Parks: Great Basin NP, Bryce Canyon, Arches NP, Zion NP, Capitol Reef NP and Monument Valley that would embellish an eclipse expedition.

Monument Valley, UT.

The American Southwest: Colorado, Arizona, New Mexico

Count the number of world-class observatories in Arizona and New Mexico and you will be reassured that you have reached the best eclipse-watching weather along the track. The centreline graph of cloud amount shows a low-amplitude up-and-down pattern that averages around 28 percent mean cloudiness. The map of Figure 9 shows a mottled blue pattern, indicating that the sunny character of the climate extends across the width and length of the eclipse path in these states.

Map of eclipse track in Four Corners area
Figure 8: Topographical map of Four Corners area.

With such a low average monthly cloudiness, the effects of topography (Figure 8) are muted. Cloud amounts increase by about 5 percent on the smaller rises such as the San Mateo Mountains west of Albuquerque and the Chuska Mountains on the Arizona-New Mexico border south of Four Corners. These ranges stand only about 1000 m above the surrounding plateau.

Figure 9. Average October cloud cover at 10:30 am derived from 19years of satellite observations (2000 -2018)

As the shadow path approaches the Texas border and moves onto the Llano Estacado, October cloud amounts begin to increase. It’s only a small upward climb, best seen in the graph of centreline cloudiness (Figure 3), but it gives a hint of things to come farther along the track. In contrast to the topography of the Colorado Plateau over Arizona and New Mexico, there is much less variation in the terrain over the Llano, so cloud cover is more a reflection of passing weather systems than topography. With an altitude of about 1100 m in New Mexico, the Llano’s plains keep much of the climatology of the Plateau.

Corona, NM.

Cloud cover in New Mexico and surrounding states comes mostly from cold fronts that sweep southward from the Great Plains and the Colorado mountains. They bring heavy cloudiness for the most part, but seldom cover the whole of the track while passing through Arizona and New Mexico. Since fronts are usually well forecast several days ahead, mobility will be a game-saving advantage. Albuquerque, located at the junction of Interstates 25 and 40, would make a good home base in the days before, but if you are confident of the weather, then New Mexico and the surrounding states have marvelous attractions to frame your eclipse experience. These include Chaco Canyon National Historical Park, Mesa Verde National Park, and, of course, Roswell. For those who put their trust in omens (or beer), Corona, New Mexico lies just off the centre line (though the corona is only briefly visible to cameras during an annular eclipse).

Table 3: Climate statistics along the eclipse track in Colorado and New Mexico

In the American Southwest, winter is coming but not yet arrived in October, though there are days when it threatens. For the most part, temperatures are pleasant or a touch cool in the day and a bit above freezing at night (Table 3). Record highs and lows demand a little more attention when it comes to eclipse-day planning, as temperatures can drop to around the 5°F (-15°C) level across much of the track. Snowfall, however, is uncommon on the lower terrain (but not the mountains) and seldom amounts to much.

graph of centreline cloudiness
Figure 3 (repeated): Graph of mean October cloud cover (2000-2018) along the centreline of the eclipse track, collected during the satellite passage at 10:30 am. Units are fractional sky cover, which can be interpreted as percent cloud cover.

Texas

The lunar shadow enters Texas over the Llano Estacado (also known as the Staked Plains), crosses a steadily descending topography onto the Edwards Plateau, and then drops over the Balcones Fault onto the Texas Coastal Plain (Figure 10). Across the Llano and the Edwards Plateau there is little in the topography to distinguish the Llano from the Plateau (though there is a considerable difference in the geology). Most important is their elevations, ranging from 1100 m in the northwest to around 600 m in the southeast, a level that impedes the influx of low-level moisture from the Gulf of Mexico into the region. As the track drops over the Balcones Fault, the track is exposed to the humidity of Gulf weather.

Map of the eclipse track through Texas
Figure 10: Topographical map along the eclipse track in Texas.

The graph of centreline cloudiness (Figure 3, repeated above) shows a steady increase in cloud cover, rising about 10 percent (from 0.33 to 0.44) in the trek to lower elevations across the plateaus. The centreline is not representative of the whole of the track, however, as Figure 11 shows that the cloud cover increases much more quickly over the south side of the path. Beyond Scheffield, the south limit “falls off” the edge of the Edwards Plateau, moving across the lower terrain of the Rio Grande Valley. The valley acts as a funnel, allowing Gulf moisture to move inland to Del Rio and beyond where it contributes to the higher cloudiness on the south side of the track. On the north side of the centreline, cloud frequencies are similar to those along the centreline, but there is more variability from place to place than was the case over the Llano Estacado.

Cloud map for Texas
Figure 11: Average (2000-2018) October cloud cover at 10:30 am over Texas.

About half-way across Texas, before Hondo, there is an abrupt 7 percent upward jump in cloud cover on the centreline graph. This jump marks the transition from the Edwards Plateau to the lowland that surround the Gulf of Mexico. The lowlands are embedded in higher humidities than on the plateaus, so there is a corresponding increase in cloudiness. In some places, average cloud amounts reach 60 percent just before the coast, though not on the centreline.

Most cloudiness in the higher terrain in Texas comes from passing cold fronts and the occasional low-pressure system that develops over the state. October is beginning the transition to winter and these systems are becoming more common as cold weather approaches.

The availability of moisture on the Gulf Plain and the lower parts of the Edwards Plateau has a significant impact on the prospects for eclipse day. The region has high frequency of overnight and morning fog that may persist as an opaque layer of stratus through the day or turn into a broken layer of cumulus clouds. Fog also forms on the Edwards Plateau, especially in the lower, eastern parts, but this fog burns off in the morning, though it may still be lingering at the time of the eclipse.

Immediately offshore, October cloudiness drops below 40 percent, indicating that a significant part of the cloud found onshore is convective, created by daytime heating of the land. Satellite observations show that in mildly convective days, a sea-breeze circulation set up that brings cool Gulf air inland where it inhibits convection. This leaves an area of clear skies along the coast for a few kilometres inland and gives the coast a much lower average cloudiness. Inland from Corpus Christi, average cloud is measured at 0.57 (57 percent) while offshore it is 20 percent lower. The Gulf coast of Texas has many inlets (such as Corpus Christi Bay) that spread the cooling effect inland, but the strongest impact will be on the coast itself. The cooling associated with the approach of the lunar shadow will also help to suppress small convective clouds.

Climate table for Texas
Table 4: Climate statistics for selected sites along the eclipse track in Texas.

There is a gamble in staying near the coast however – if the day is particularly unstable, then the collision of the incoming sea breeze with unstable air over land will build deep convective clouds that may grow and spread high-level cloud back over the beaches. This is usually an afternoon phenomenon, but cannot be discounted during the morning arrival of the eclipse. The development of convection will depend on the instability of the day, which is well forecast by conventional weather models (just ask a storm-chaser).

Severe Weather

Severe convective weather with hail, wind, rain and tornadoes can come at any time of the year in the United States, but October is well out of the main storm season. The lightning strike map of Figure 12 shows that storms are mostly unknown through Oregon, California, and much of Nevada during the month. On reaching Four Corners, however, the frequency of strikes increases sharply, reaching a peak over Texas. Most of these are not severe storms and their main impact would lie in the large amounts of high- and mid-level cloud that they throw up on their approach. Severe weather, including tornadoes, is still possible despite the late season, but the frequency is lower along the path of the eclipse than in other parts of Texas (Figure 13).

Lightning map
Figure 12: October lightning strike distribution 2000-2012.

Storm chasers are usually more interested in finding than avoiding thunderstorms, but some of the techniques they use could be applied to eclipse chasing. In most cases, it is best to get behind the storms, as high-level cloud spreads a long way downwind (which would probably be toward the northeast on a typical storm day) and is difficult to predict, even over a short time. Avoid driving through the storms (called “core punching”) to get to the back side—the weather within the storms may have very unpleasant surprises.

October tornadoes
Figure 13: October tornado reports 1950-2018.

Tornadoes are small circulations; tropical storms are big. In the years since 1980, 13 tropical storms have made landfall in Texas in October. Ten of these brought heavy rain from a past, decaying hurricane that moved across Mexico from the Eastern Pacific; the remainder were Gulf of Mexico-born storms. There is nothing to be done when they approach other than to go elsewhere. Fortunately, the frequency of tropical storms in October is one-half of that at the peak in early September.

Central America and the Caribbean

The influence of the subtropical anticyclones is felt most strongly over the waters of the Gulf of Mexico, where sinking air warms and dries the atmosphere. The warm air aloft prevents the growth of deep convective clouds, and daily weather consists mostly of small cumulus clouds that struggle into the air. Heavier cloud comes with the occasional North American cold front that manages to push into the Gulf and from the occasional tropical storm as it moves through the area.

The stable atmospheric conditions brought on by the anticyclones persists across the Yucatan Peninsula where the eclipse track comes back onto land. Daily weather here consists mostly of the same cumulus buildups as over the Gulf but cloud amounts are heavier because of the greater solar heating of the ground and in the afternoons, the convective clouds may reach grow high enough to bring rain. Cloudiness gradually increases from north to south across the Yucatan, from around 60 percent on the Gulf coast in Mexico (Figure 14) to around 80 percent along the beaches of Belize. The best chances lie on along the coast near Campeche where sea-breeze winds give a sunnier climate than elsewhere over the peninsula. The Aqua satellite measures an average cloudiness of about 48 percent on this coast.

Cloud-cover graph along the eclipse track in Central America.
Figure 14: Graph of average October cloud cover along the centreline of the eclipse track through Central America at 1:30 pm. Data are median cloud fractions and may be treated as percent cloud cover.

As the shadow path leaves the Yucatan and crosses the Gulf of Honduras, cloud amounts decline over water, but then comes back with very discouraging values over Honduras and Nicaragua and eclipse travellers must be careful to select their viewing site. The anticyclones are losing their influence here and cloud suppression is not as effective as farther north. Nevertheless, there are one or two sunny refuges to exploit, most particularly on the north coast of Honduras, at La Ceiba (Figure 15).

Figure 15: Average October median cloud cover (2000 – 2019) at 1:30 pm along the eclipse track over Central America.

The explanation for this sunny oasis lies in the nature of the terrain in Honduras. La Cieba is backed by the Cordillera Nombre del Dios that lie a short distance (12 to 24 km) inland, reaching altitude of over 2300 m. The prevailing southeast trade winds will be blocked by the cordillera, leading to a downslope chinook-like clearing on the coastal side (Figure 16). Examination of daily satellite images indicates that the effect is most pronounced in the morning. Because morning skies tend to have only small cumulus (unless a larger weather system is present), it also suggests that the incoming eclipse shadow will be effective at clearing small convective clouds on the morning of October 14, making La Cieba an attractive destination for viewing the eclipse.

Figure 16: Central American topography along the eclipse track.

Just offshore from La Cieba lies the islands of Utila and Roatan; Utila is on the centreline; Roatan stretches from the north limit to about half-way to the centre. Both are very popular tourist destinations, but lie just a bit too far offshore to participate in the clearer skies along the coast of the mainland. Especially on the larger Roatan, cloud cover is generated by the heating of the ground, so eclipse watching there should be enjoyed on the windward side of the islands, where cool air comes ashore.

From La Ceiba to the Gulf of Panama, the eclipse track passes into a region of very high cloudiness. Over Honduras and Nicaragua, the cloud cover comes from solar heating, which comes on quickly in the morning and builds a heavy afternoon cloud. The eclipse is now into the afternoon here, and prospects are not encouraging. There are a few refuges—the centreline graph shows a decline in cloud cover south of Catacamas, Nicaragua, where the track descends from a 2000-m mountain range to lower elevations, but the improvement still leaves average cloud amounts above 70 percent. There is another overwater drop offshore from Bluefields, which is possibly best sampled at Limon, barely inside the eclipse track in Costa Rica, but one cannot be encouraged by a “fall” to 80 percent average monthly cloud cover.

The heavy cloud from Nicaragua southward comes as the eclipse track moves toward the Intertropical Convergence Zone. The beneficial effects of the Atlantic anticyclone are completely gone at this latitude and the Central American peninsula is now dominated by low-pressure systems. When the Moon’s shadow reaches Panama’s Peninsula de Azuero and moves out over the Gulf of Panama, it encounters the full effects of the ITCZ and cloud cover settles at 80 to 95 percent. In 20 years of satellite imagery for eclipse day, Panama enjoyed only one where eclipse watching would have been easy and four where a lucky location could have viewed the event.

Table 5: Climate statistics over Central America at selected sites along the eclipse track.

It is easy to become overly pessimistic about eclipse-viewing prospects when looking at the satellite cloud-cover maps, but, especially in higher cloud locations, the satellite algorithms seem too pessimistic. This might be due to the presence of large amounts of cirrus cloud which appear opaque to the satellite but are transparent enough to pass sunshine. Some evidence of this comes from the sunshine statistics (Table 5) for stations in Central America where even in Panama, the average October day experiences sunshine during 35% of the hours.

Colombia and Brazil

The South American portion of the eclipse begins in heavy weather and ends in sunshine with modest ups and downs in cloud cover along the way. In October, the ITCZ has a strong footprint over the Pacific, but the convergence zone breaks down over the continent, becoming instead a broad area of convective cloudiness that covers almost all of the eclipse track. It is only at the end of the track, over Brazil’s Northeast Region, that convective weather loosens its hold on the climate and fine weather returns.

Colombia

Colombia enjoys the unlucky distinction of having three of the ten cloudiest cities in the world—a not surprising declaration given the unfortunate climatic factors that converge on the country. A tropical latitude, a humid Pacific Ocean, prevailing westerly winds that bring marine moisture onto land, the nearby ITCZ, and a towering chain of mountains that traps jungle moisture over its interior plains would seem to discourage the most optimistic of eclipse seekers, particularly as October lies at the peak of one of its two rainy seasons. In spite of this, there are a few less cloudy hideaways where the chances of catching the Sun are encouraging.

Figure 17: Topographic map along the eclipse track over Colombia and Brazil.

Along the coast of South America, prevailing winds blow from the south, following the coast, until they reach the latitude of Colombia, where they turn to the east to intercept the coast. After a short 100 km passage across a low coastal plain, the Pacific flow must climb the 3,000 metre heights of the Cordillera Occidental (Figure 17), the first of three mountain ranges in Columbia that make up the Andes Chain. The Occidentals are followed by the higher Cordillera Central and then the Cordillera Oriental (Figure 18). Between the mountain chains are two deep valleys. The Rio Cauco Valley is 2,000 m lower than the peaks of the Cordillera Occidental and the Rio Magdalena Valley drops 3,000 m from the peaks of the Cordillera Central. On the east side of the Magdalena Valley, the peaks of the Cordillera Oriental provide another towering barrier before the Moon’s shadow drops down to the Amazon Basin, a low-elevation jungle landscape where elevations barely exceed 100 m.

Eclipse topography over Colombia
Figure 18: Cross section of the terrain along the eclipse centreline over Colombia.

When air rises to go over a mountain barrier, it cools and its relative humidity climbs, leading to saturation of the air mass and the formation of cloud and precipitation. On the Pacific side of the Cordillera Occidental, annual rainfall along the coast exceeds 6,000 mm at Buenaventura and a touch under 9,000 mm just beyond the north limit at Quibdo. In the inter-mountain valleys along the Rio Cauco and Rio Magdalena, air arrives after descending from the heights, warming and drying along the way. The descending air eats away at the cloud cover, giving the valleys a lower cloudiness for the month than the coastal regions or the mountain slopes. The drying effects of the intermountain valleys is mirrored in the precipitation amounts (Table 6) reported at Cali and Neiva compared to that at Buenaventura.

Table 6: Climate statistics at selected stations along the eclipse track over Colombia and Brazil.

In the cloud-cover map of Figure 20 and the cloud graph of Figure 19, the pattern of cloudiness is tightly tied to the terrain, with greater amounts on the slopes and mountaintops and lower values in the intermountain valleys. There is a hint of lower amounts along the coastline, especially south of the centre line. East of the Andes, over the lowland jungles of the Amazon Basin, cloud cover is high but relatively even, as there is little topography to break up the landscape.

Figure 19: Median cloud cover along the eclipse centreline at 1:30 pm derived from 20 years of satellite data.

Most of the cloud cover in Colombia is convectively generated, but the cloud buildups are early risers, not waiting until the heat of the afternoon to blossom, though they do grow larger toward the latter part of the day. Over the Andes, the convection begins after sunrise on slopes, leaving the valleys open until later in the day. Even in the cloudiest days, the valley bottoms tend to have some openings, suggesting that a mobile eclipse watcher would have a good chance of finding a place to watch the Sun, particularly as it is relatively high in the sky (60°). The task of finding a suitable opening is helped by the direction of the major highways, which run north-and-south across the track along the valleys rather than across the mountains. In 19 years of satellite coverage, an observer at Neiva would have been able to see the eclipse with no or a small amount of movement on 13 occasions, four of which would be through thin cirrus. Table 6 shows that Neiva sees an average of 46% of the daily sunshine in the month, which fits well with the results of the daily visual assessment of the satellite images, assuming that a cirrus-filled sky would be measured as “cloudy” by the sunshine recorders.

Figure 20: Average October cloud cover acquired from 20 years of satellite imagery (2000 -19) over Colombia and Brazil.

Daily satellite images (as opposed to monthly cloud averages) and the sunshine measurements in the intermountain valleys give much more optimistic evaluations of the amount of cloud cover than the automated measurements from the Aqua and Terra satellites. The cloud amounts determined by the computer algorithms that provide the statistics for Figure 19’s graph are determined across an image element that is 10 km on a side, a dimension that often spans the valley. The algorithm determines whether the pixel is cloudy or not based on the radiance values from several wavelength bands in the visible and infrared. There is no “partly cloudy.” (The algorithm actually distinguishes between “cloudy” and “probably cloudy” but both are treated as cloudy.) Unfortunately for our purposes, a 10-km element is larger than the valley bottom and typically includes clouds that are on the slopes of the mountains. The average of clear and cloudy radiances of the pixel then comes out as cloudy and so we end up with a grayer climatology than is actually present. The satellite measurements of cloud cover usually work quite well, but in Colombia’s climate, where cloudiness varies over small distances, the fine details get smeared out.

There are errors in all methods of cloud cover measurement and so the cloud maps should be taken as an indication of relative cloudiness rather than absolute amounts. In other words, use the maps and graph to evaluate whether one site is better than another and not what are your chances of seeing the eclipse.

About 38 km north of Neiva, very close to the centreline, is the Desierto de la Tatacoa whose name suggests a particularly dependable location, though lack of precipitation does not necessarily imply a lack of clouds. It’s a scrubby landscape with wide, open skies and, because it’s a tourist attraction, passable roads and some facilities. You may even see a few observatories in backyards along the way. Wikipedia describes it as “ideal for astronomy,” though the dry and open landscape may give rise to daytime thermals that corrode the best views of the Sun. The desert lies at the lowest point of the graph of cloudiness in Figure 19, with a satellite-derived value of 75%, among the lowest along the track in Colombia.

On the eastern side of the Andes, satellite data display a cloudiness that is comparable to that in the mountain valleys within the cordillera. There are a few locations that have slightly better cloud-cover prospects, but only by a few percentage points, but the road network is extremely limited, and most of the shadow track would only be reachable by air.

Brazil

The lunar shadow enters Brazil over the flat jungle terrain of the Amazon Basin where the climate hardly varies until the track is most of the way across the country. In October, all cloudiness comes from convective buildups, a process that is very sensitive to underlying temperatures. In Figure 20, we can see the outlines of the major rivers in the pattern of cloudiness, a sign that a significant part of the convection is low-level cumulus cloud that is suppressed over the cooler river temperatures. The effect of this over-water suppression is significant, as satellite-measured cloudiness drops by more than 20% compared to values over land. This may augur well for eclipse day, as the cooler temperatures brought on by the approaching shadow should do the same, though the effect will be small. Cloud-cover statistics measured at airports show only a small increase in cloud-free skies as night falls.

While day-to-day satellite images do show the important role played by cumulus clouds in the monthly climatology, they are a small part of the cloud cover over the Amazon region. Much more important is the higher layers of cloud cover caused by thunderstorms that deposit large amounts of moisture in the upper atmosphere. These higher clouds are responsible for most of the overlying cloudiness away from the rivers. Table 6 shows that sunshine, even in the cloudiest areas, is not entirely absent, though the percent of maximum sunshine column has values that are lower than anywhere else along the track. Humans place their cities in the bottom of valleys and so sunshine measurements within mountains will reflect the best of conditions.

As the eclipse track proceeds eastward, cloud cover grows, from around 75% to 90% as the shadow path moves onto the Brazilian Highlands. The increase may or may not be related to the terrain, as the highest cloudiness comes where the topography drops downward around Parauapebas and Araguaina. The descent is small, only about 200 m.

Further along the track, terrain does seem to play an important role, and here Brazil brings a bit of a surprise. Beyond Araguaina, October cloudiness begins to decline and it isn’t a small drop. As we can see in the graph of centreline cloudiness (Figure 19), cloud amounts fall from over 90% to just above 30%, a feature that gives an eclipse climate that is as good as any in the United States. This is the sertão, a region of low rainfall that sits overtop the eastern part of the Brazilian Highlands. The lack of rain and cloud is explained as being due to the suppression of convection by the nearby South Atlantic Ocean, but this is not completely satisfactory, as the lower coastal plain, against the ocean, has much more cloud. The sertão’s altitude, about 1000 m above sea level and out of the largest part of low-level moisture from the ocean, may also be a factor. Almost all of the cloudiness over the sertão comes from small cumulus clouds that would likely succumb to the oncoming lunar shadow.

Measurements from stations along the track confirm the pattern of cloudiness portrayed by the satellite images. Tefe and Manaus 40 to 45 percent measurement contrasts with the 70 to 77 percent values at Natal and João Pessoa. Sunshine percentages reach almost to 80 percent at Natal and it’s probably higher in the middle of the sertão. The same pattern is also reflected in the October precipitation figures. At Natal, October is the driest month of the year, so the eclipse comes at an opportune time. These figures all combine to acknowledge the sertão, especially around Iguatu and Caicó, as South America’s prime eclipse-viewing spot, even though, at central eclipse, the Sun is only 8° high and sets before fourth contact.

As the eclipse track drops down to the coast, the terrain declines about 600 m to create an ocean-facing bulwark that intercepts the prevailing winds and forces the air upward. This forced rise causes a sharp increase in cloudiness over the last 100 to 200 km of the land-based part of the eclipse track. While the cloud is still largely convective, it is deeper and heavier than over the sertão. Right at the coast, sea-breeze winds suppress the convective clouds for about 5 to 10 km inland, creating another small refuge where average cloudiness is about 45%. The sea breeze is most effective from João Pessoa to Natal; the south limit sees less advantage.

Summary

The range of climates along this eclipse track presents both opportunity and challenge. For much of it, travel is difficult and the eclipse spectator must take whatever the day brings. It is also unfortunate that the places with the greatest cloud cover are also those with limits to quick movement. The two regions with the best weather, the U.S. Midwest and the Brazilian sertão,  have decent highways on which to move—an embarrassment of riches.

Supplementary satellite images of weather patterns along the eclipse track.

Click on the box above to go to a selection of satellite images from past October 14s.

Updated: April 2020