Last year, I wrote a long post on this blog giving the most general intuitive explanations possible for the existence of different climate types, based on different configurations of land & ocean at different latitudes, and used that to explain broad & subtle features of actual climates of most locations. As I alluded to in that post, I had a lingering question of why, at tropical latitudes, there are so many east & poleward coasts that have dry seasons despite those coasts getting ordinary easterly (sometimes with a poleward originating component too) tradewinds throughout the year largely perpendicular to the coast. My confusion is because at those latitudes, the water & land temperatures are warm enough even in the winter half of the year to suggest that humid air from over the ocean could unstably rise above air over land & lead to precipitation. Examples include but are not limited to the north (poleward)/east coast of Central America, the north (poleward) & east coasts of islands in the Caribbean, some parts of the coast of Brazil, the east coast of the southern part of India, and the east coast of the northeastern part of the mainland of Australia. I also had a question about why the north coast of Egypt did not get precipitation in the summer half of the year despite getting northerly onshore winds from over the Mediterranean Sea, which is warm enough that despite the air over land becoming considerably hotter then, humid air from over the Mediterranean Sea could in principle rise above dry air over land, leading to instability and therefore precipitation. Additionally, I had a question about why the subtropical ridge over each ocean (more prominently over the Atlantic & Pacific Oceans in the northern hemisphere) is more poleward along the eastern edge of the ocean even though the western edge has warmer currents (suggesting lower pressure along the western edge than along the eastern edge along the same latitudes).
I could not think of satisfactory answers until very recently. Ironically, although that linked post from last year referred to concepts beyond surface-level wind patterns & air pressures, these answers depend mostly on surface-level wind patterns along with the knowledge of what different surface-level wind patterns imply for vertical air flow. The sources that I used are many relevant pages from Wikipedia and the Columbia University interactive maps of mean monthly wind velocities; unfortunately, the latter resource will be shut down in 2026 April due to funding difficulties. Again, I am not a trained climatologist or meteorologist; I can't guarantee that this information is accurate, and I can only say that my intuitions seem through my limited understanding to align with superficial aspects of more detailed explanations. Follow the jump to see these answers.
Convergence and Divergence
Along the surface of the Earth or at any fixed elevation above sea level, the horizontal components of winds may converge or diverge. The following explanations are very rough, and this resource from the National Weather Service makes clear how vertical motion can be nullified by one form of convergence occurring simultaneously with the other form of divergence.
Convergence can be seen on a map of wind velocities by seeing the magnitude of the velocity decrease along a streamline or by seeing nearby streamlines coming closer together. In either case, the mass balance may be maintained by some of the air rising to higher elevations. If that air is humid, likely from having come from over an ocean or large sea, then precipitation is more likely.
Divergence can be seen on a map of wind velocities by seeing the magnitude of the velocity increase along a streamline or by seeing nearby streamlines separating from each other. In either case, the mass balance may be maintained by some of the air falling from higher elevations. That air will always be relatively dry, making precipitation less likely.
Previously, I could only identify obvious areas of convergence or divergence corresponding to systems of low or high pressure; these were identifiable by counterclockwise or clockwise flows of air around a point or curve (with the direction depending on the hemisphere & type of pressure system). I have now improved my ability to carefully identify convergence & divergence on smaller length scales, even when those convergent or divergent wind patterns do not create significant systems of low or high pressure, respectively, measurable in monthly averages.
Explanations of Precipitation Patterns at Tropical Latitudes
At every tropical location, even those along east coasts under the influence of ordinary easterly tradewinds largely perpendicular to those east coasts through the year, the dry season occurs when the tradewinds diverge in speed or direction (requiring inflow from above, implying high pressure/stable inversion), while the wet season occurs when the tradewinds converge in speed or direction (requiring outflow to above, implying low pressure/instability).
In many cases, the explanation is particularly simple. In the summer half of the year, the ITCZ is closer than the subtropical ridge to the location in question, so the tradewinds are decelerating along streamlines and therefore converging in that location, leading to more precipitation. By contrast, in the winter half of the year, the opposite occurs in every respect (and even if mountains are present, those mountains are usually not tall enough to force convergence to a degree that would undo the divergence from accelerating tradewinds, as is the case in Chennai in India windward of the Eastern Ghats in the winter season or of much of Cuba during the winter half of the year).
This phenomenon of changing wind speeds along streamlines could be illustrated by drawing curves connecting the points where the magnitude of velocity along a streamline is locally maximized and saying, to a decent approximation, that locations between that curve and the ITCZ experience convergence and therefore more precipitation whereas locations on the other side of that curve (away from the ITCZ) experience divergence and therefore less precipitation. Those curves are similar but not identical to velocity scalar potential contours, as the existence of rotational flow around systems of high or low pressure means that the velocity field in 2 dimensions has a contribution from the gradient of a scalar potential and a contribution from the curl of a vector potential (which points vertically, out of the surface, and whose scalar component in that direction is the stream function). Mathematically, if the velocity vector field \( \vec{u}(\vec{x}) \) is defined only on the surface and time dependence of the velocity vector field is ignored, then \( \vec{u}(\vec{x}) = -\nabla \phi(\vec{x}) + \nabla \times (\psi(\vec{x})\vec{e}_{z}) \). The source of the scalar potential \( \phi \) would be the local density deviation from the mean, analogous to stationary electric charges effecting irrotational static electric fields. The source of the stream function \( \psi \) would be vertical flows of air modulated by Earth's gravity for vertical flows & the Coriolis effect for rotation, analogous to steady electric currents effecting incompressible static magnetic fields rotating in planes perpendicular to the currents.
However, the broader explanation that includes convergence & divergence in direction too still holds even if convergence or divergence in speed are not as easy to observe. For example, along the north coast of Brazil (which is very close to the equator), the tradewinds from both hemispheres converge during the first half of the year but diverge during the second half of the year between going toward the ITCZ farther north over land in the northern hemisphere versus going to the ITCZ farther into the interior of Brazil (and even that interior part of Brazil, which is in the southern hemisphere, has a dry winter season despite supporting the ITCZ then because the surrounding oceans are cooler then than during the summer half of the year). Meanwhile, along the east coast of Brazil & the eastern part of the south coast of Brazil (which are firmly in the southern hemisphere), the divergence corresponding to the dry season is for tradewinds splitting between going toward the interiors of Brazil near the equator versus of Argentina in middle latitudes.
This point about convergence & divergence is an additional explanation, compatible with explanations based on upwelling caused by the reversed southwesterly tradewinds in the summer half of the year along with the tradewinds through the year being parallel (rather than perpendicular) to the coast, for why Somalia is a desert. There is relatively more divergence in speed & direction along the coast of Somalia in the summer half of the year with the reversed southwesterly tradewinds than in the winter half of the year with the ordinary northeasterly tradewinds, and the divergence in the summer half of the year is reinforced by upwelling but in the winter half of the year is attenuated by downwelling.
Explanations of Precipitation Patterns along the North Coast of Egypt
This point about convergence & divergence even explains why the north coast of Egypt (which is at middle latitudes) as well as both coasts of the Red Sea get precipitation during the winter half of the year but not during the summer half of the year, despite the north coast of Egypt getting northerly tradewinds from over the warm Mediterranean Sea. Greater proximity to the subtropical ridge during the summer half of the year means that those areas experience accelerating & directionally diverging winds; this holds even though the surface pressure there is considerably lower than in the middle of the subtropical ridge over the Atlantic Ocean in the northern hemisphere in the summer half of the year. Additionally, the north coast of Egypt gets little precipitation even in the winter half of the year, making Egypt a desert, because even during the winter season, the prevailing westerlies are parallel to the coast and exhibit noticeable directional divergence despite exhibiting slight speed convergence.
Explanations of Precipitation Patterns along West Coasts at Poleward Middle Latitudes
Even far outside of tropical latitudes, this point about convergence & divergence is an additional explanation for why the west coasts of North & South America at poleward middle latitudes as well as the west coast of Norway get less precipitation on average per month in the spring season than in the later part of the summer season & in the fall season, though precipitation is very high through the year. This explanation is compatible with explanations based on the land still warming enough for cool moist air from over the ocean having a greater chance than in the winter half of the year to stabilize the atmosphere.
Explanations of Subtropical Ridges over Oceans Being Closer to the Pole along Eastern Edges
Convergence also explains why subtropical ridges over oceans are closer to the pole along eastern edges of those oceans. The following argument applies to the northern hemisphere, but it can be made to apply to the southern hemisphere by exchanging north & south. The Coriolis force means that tradewinds from the equatorward side of the subtropical ridge over an ocean in the northern hemisphere are northeasterly and turn clockwise to become easterly, so they converge into the ITCZ over the ocean close to the equator only by having the eastern edge of the subtropical ridge being further north. By contrast, if the western edge subtropical ridge were further north, the tradewinds on the equatorward side of the subtropical ridge would start out easterly and then turn away from the equator by being southeasterly, which would be a divergence that contradicts the observation of easterly convergence at the ITCZ.
This is essentially a boundary value argument from the study of differential equations. It is not the most satisfying explanation to me, because it still doesn't fundamentally explain from the perspective of the Hadley cell and the different temperatures of ocean boundary currents why air at high elevation that moves toward the pole starts to descend sooner (more equatorward) along the western edge than along the eastern edge. That said, I am willing to accept this explanation provisionally.
Concluding Remarks
In the linked post from last year, I thought that I wouldn't write more posts about climate explanations as I was largely satisfied with my broad understanding of why different parts of different continents have the climates that they do. Perhaps I was deluding myself because even that post clearly showed that I had unanswered questions, and I felt compelled at a few points over the rest of the year to try to answer those questions. Additionally, I did ultimately write a few more posts about climates in locations other than east coasts at tropical latitudes.
Now that I have answered those questions, I genuinely feel like I don't have any more unanswered questions about why different locations have the climates that they do. For this reason, I genuinely believe that I will be much less likely to write posts on this blog about climate explanations (though I don't want to completely foreclose the possibility). Additionally, I now realize that a map of locations by climate type is made not only redundant by the broad understanding that I developed when I wrote the linked post from last year (as I discussed at the end of that post) but also futile by the fact that surface winds at a given location follow streamlines that are affected by phenomena many hundreds or thousands of miles away (as can be seen by winds along the east & south coasts of Brazil diverging in the summer half of the year, with one branch going as far as the interior of Argentina to a seasonal system of low pressure there), so the climate of a location cannot be explained only by the location's latitude, position relative to a coast, closest ocean water temperatures, and other purely local geographic features. This truly reinforces the idea from the end of the linked post from last year that my continued compilation of such a map is useful for building my skills with GIS and for giving a basic but incomplete understanding of climate types by location.
