Seasonality of the Catalan Triangular Trade

Seasonal variation in supply and demand for tasajo must affected the timing of the triangular Catalan trade with the Rio de la Plata and Cuba. Cuban demand for tasajo increased during the cane harvest from January through June, as plantation owners bought additional enslaved Africans for the grueling work, which killed many. Meanwhile, tasajo production took place during the nine months of October through June, the southern hemisphere’s spring, summer, and fall. During the winter, in contrast, the cattle weighed less and the air temperature was too cold to dry tasajo. To maximize the price differential between the purchase of the tasajo and its sale, then, voyages departed from the Río de la Plata during the season of tasajo production in time to arrive during the season of peak demand in Cuba.

The export statistics for the port of Buenos Aires confirm that seasonal rhythm (after fig. 6.12 in my book Black Ranching Frontiers, Yale,2012). Shipments of tasajo through the port of Buenos Aires began as soon as tasajo became available in the spring (October and, even, September) to arrive in Cuba in November and December in anticipation of the beginning of harvest season. That volume of shipments continued until about May, with those departing at that time still able to arrive in Cuba before the end of the cane harvest. July and August marked the nadir of tasajo shipments northward due to a lack of both supply and demand.

A temporal analysis run in ArcMap, shared using the video below, clearly shows the seasonal variation in the triangular trade, even with only fourteen voyages. I achieved the visualization by changing the year of each voyage to a single year, setting both the time interval and time window to one week, and creating a new point feature in the South Atlantic labeled with the month field. The daily-position symbols are triangles, with a week’s worth showing at a time. Their colors represent the direction of the voyage segment, based on their destination fields: green for northbound, orange for southbound, yellow for eastbound, and purple for the single westbound segment.

The visualization shows that in September, as cattle began to fatten, temperatures started to increase, and tasajo production kicked off, shipments began to leave the Rio de la Plata. They arrived in Cuba by November or December in anticipation of the cane harvest that would begin after Christmas. The shipments continued until tasajo production and the cane harvest both wrapped up in June. Focusing on the Caribbean during July and August reveals the lack of northbound (green) segments those two months.

Animated Acancia-Hurricane Interactions

Despite the lack of support for temporal GIS in the free, non-subscription version of ArcGIS Online, ArcMap does support the creation of videos that can be uploaded to free blogs like this one or to video sites like YouTube. To create a video, users zoom to the desired area of the GIS, such as the North Atlantic, and then open the Time Slider window to set controls such as time extent, interval, and window. They then use the Animation toolbar to select Create Time Animation. The Animation Controls and Animation Manager, available through the same toolbar, allow adjustment of variables like the length of the animation in real time and playback time. Export Animation opens a dialog box to set the video format, choose compression options, and save the video.

I exported such a video focused on the North Atlantic with the route and daily position symbols for the Acancia and the HURDAT layers turned on using August 1 through September 5, 1893 as the time extent and 1 day as the time interval and window. The result was AcanciaAug1893.avi, which shows the vessel's day-by-day progress along its route line with hurricanes tracking nearby. Since each 24-hour time window has two noontime position markers, the current and previous day's noontime positions are visible on each frame, allowing users to visualize the daily progress of the Acancia along its route. Similarly, since each hurricane has several positions per day in HURDAT, at 6 hour intervals, the time settings allow users to visualize the evolution of the hurricane tracks through time and space even without the equivalent of a hurricane route line.

AcanciaAug1893.avi was 950 MB in size, however, far in excess of the 100 MB upload limit for this blog. I therefore opened it in Windows Live Movie Maker, included in the most recent versions of MS Windows and available from Microsoft as a free download, to reduce the file size. After adding some title frames, I saved it as AcanciaAug1893.wmv, only 8.3 MB, and uploaded it to this blog.

The result allows readers to view an animated version of the crossing of the North Atlantic by the Acancia in August of 1893 that is fairly identical to what they would see using the time slider in ArcMap. The resolution is lower and the user cannot pan, zoom, or control the time animation as fully as in ArcMap, but the video nonetheless visualizes vessel-hurricane interactions. As the Acancia passed north of Bermuda into the mid Atlantic with its cargo of aguardiente, a tropical storm formed in its wake and tracked northward. To the south, several hurricanes arced westward from the Cape Verdes toward the Caribbean, Florida, and the Atlantic Seaboard. As the Acancia turned southward toward its destination, Montevideo, it sailed directly across that cordon of hurricanes, and on August 26th at noon came within 50 miles (80 km) of the eye of a category 2. After an eastward excursion from its southward course, the Acancia continued southward toward Montevideo.

Since users have much more limited control to pan, zoom, and so on than if ArcMap could be time enabled on ArcGIS Online, I created another video that focuses more closely on the mid-Atlantic encounter between the hurricane, the Acancia, and its crew in late August 1893.

In future posts, I will explore other ways to share the temporal aspects of the GIS on the Web.

The Acancia meets Hurricane 7 of 1893

To more closely examine encounters of particular vessels and hurricanes would benefit greatly with being able to run the voyage and hurricane layers through time, as I can with my ArcMap version of the GIS. While I cannot upload time enabled files to the free version of ArcGIS Online, NOAA has published HURDAT on ArcGIS Online as a map service and as a Web map. The map service is served from the following URL: http://maps4.arcgisonline.com/ArcGIS/rest/services/A-16/NOAA_Hurricane_Tracks-Temporal/MapServer, and anyone can therefore add it as a layer to their Web map using the free ArcGIS Online viewer and display the data through time using the temporal function.

That is what I have done experimentally in the below Web map, reducing the voyages to that of the Acancia in order to simplfy the map during the trial. To add such layers, select Add Data, Add Layer from the Web, and then type the URL for the service into the dialog box. After being added to the Web map, the time control, or "time slider,"  for the NOAA layer appears at the bottom of the map frame. In the version embedded in this blog, select the time button and the time slider will appear as a separate window.

View Larger Map

The problem is that it does not work well because of the enormous amount of data. The HURDAT layer in my ArcMap GIS, which covers the North Atlantic for 1851-1900, has 9366 records. The NOAA map service includes Pacific and Atlantic storms, 1851-2011. When running the time sequence, the rendering takes so long that blobs of hurricane tracks, indistinguishable from one another, jerk across the screen.

I tried filtering the layer to reduce the amount of data being displayed, which is also possible in ArcGIS Online for layers from a map service. I selected each sub-layer in turn, selected filter, and set the Hurricane Season field to 1893, which was the year the Acancia was crossing the North Atlantic in August. Then I set the controls for the time slider to display only 1893 and do it in 365 equal intervals to get a daily interval. Even at an extent, or "time window," of one year, however, the time function does not work well enough to use because the symbols take so long to render that they remain essentially static.

Even the data in the pop-up boxes for the hurricane symbols take a long time to arrive from the server compared to those for the Acancia. Nonetheless, the pop-ups do reveal that the Acancia, shortly after turning southward, had a near encounter with a category 2 hurricane in mid Atlantic on August 26, 1893.

In the next post, I will try another strategy to visualize discrete hurricane-vessel encounters.

The Tasajo Voyages and Hurricane Season

One of the aspects of Atlantic commodity networks like the one that involved Catalan vessels and tasajo that interests me are their seasonal patterns. In what months did the various legs, or segments, of the voyage typically occur? What did seasonal timing have to do with the production schedules of the commodities, such as the months in which various crops were harvested? What did seasonal timing have to do with sailing hazards, such as the hurricanes that characteristically occur in the North Atlantic between early June and late November, with mid August through mid October marking the peak of hurricane season?

Because I am using the free, non-subscription version of ArcGIS Online, at least so far, I cannot implement the time function of my ArcMap version as a Web map. Nor can I do queries and filters with the Web map like I can on my computer in ArcMap. But another way to analyse such patterns in time and space involves applying symbols to represent months or seasons instead of cargoes.

To do that with the objective of visualizing patterns relative hurricane season, I saved the Web map from the previous post (Catalan Tasajo Voyages, 1837-1900) as Catalan Tasajo Voyages: Seasonality. Then I applied new symbols based on the Month field instead of the Cargoes field. For peak hurricane season (roughly August through October, or Months 8-10) I applied red. For the tail ends of hurricane season (June, July, and November, or Months 6, 7, and 11) I applied yellow. And for the rest of the year (Months 12 and 1-5, or December through May) I applied blue.

View Larger Map

The overall pattern indicates that most voyages crossed the North Atlantic during low hurricane season (December-May) or during the tails of the hurricane season (June-July and November). Still, out of fourteen voyages, five have red position symbols in the northeastern quadrant of the during peak hurricane season. By zooming in on the northeastern quadrant of the Atlantic and using the pop-ups to get more information on those five voyages, they become distinct. Two, the Romantico in 1864 and the Prudente in 1884, Left Montevideo bound for Cuba with tasajo in August and September respectively, entered the zone of hurricanes in October, and reached Cuba in November, at the tail end of hurricane season. Two others left Cuba bound for Europe during hurricane season: another voyage of the Romantico, in 1863, with a cargo of sugar and the Pepe in 1851. The Pepe crossed the North Atlantic in late October and November, during the tail of the hurricane season. The Romantico left imprudently in late July and crossed during the August peak of hurricane season. And one, the Acancia, left Cuba bound for Montevideo with a cargo of aguardiente in early August of 1893, sailing first eastward and then southward across the hurricane zone during the peak of the season.

The survival of the logbooks testifies to the lack of a fatal encounter with a hurricane on any of those imprudent voyages, but adding a hurricane layer to the Web map to examine any close encounters between vessel and hurricane will nonetheless be the task of the next post.

Catalan Tasajo Voyages, 1837-1900

This Web map shows the routes of fourteen voyages by twelve Catalan vessels from the mid to late nineteenth century. For the most part, they carried manufactured goods, wine, and other Catalan products from Barcelona to Buenos Aires or Montevideo. There they loaded tasajo, a type of salt beef used as food for enslaved Africans on sugar plantations in Cuba. Once in Cuba, they either loaded sugar or aguardiente, a type of cane alcohol, and returned to Barcelona; or they proceeded in ballast for ports in the US South, like Charleston, and loaded cotton for the textile mills of Barcelona. The GIS includes several variations on that basic trade pattern. See my 2010 article on The Hispanic Atlantic's Tasajo Trail in the Latin American Research Review for more details.

For method and data, see the first four posts of May 2013. The main difference with the previously posted Web map of Dutch slaving voyages is that I did not delete as many of the fields in the Catalan layers of the GIS before uploading them. That slowed the upload process a bit, even causing timeouts on a few occasions because I am working from my home office on a relatively slow connection at the moment) but it transferred more of the data in the GIS to the Web map. The fields I did delete were related to weather observations, included in the GIS because of all the weather data in the CLIWOC database but mainly empty fields regarding the Catalan voyages because when I transcribed those logbooks I had limited time, did not need the detailed weather observations for the research project I was working on, and so focused on the latitude, longitude, and cargo data.

I will use this Web map as the basis for others in the next several posts.

View Larger Map

The project GIS and website are © 2013 by Andrew Sluyter but open source and licensed through the Creative Commons as attribution-noncommercial 3.0, which allows others to use the data and programming to produce non-commercial derivative products. I collected the data in four archives in Spain while funded by a 2009-10 Louisiana Board of Regents ATLAS grant: Arxiu Històric Municipal del Masnou, Arxiu Històric Municipal de Sitges, Biblioteca de Catalunya, and Museu Marítim De Barcelona.I processed the data and built the GIS while a 2012-13 Digital Innovation Fellow of the American Council of Learned Societies. As with the other Web maps on this blog, anyone is free to use it as a basis for new, non-commercial, scholarly and creative products as long as each such GIS, website, and/or publication contains a statement that acknowledges Andrew Sluyter as this GIS's creator.