The SeedMapper

In this post we present SeedMapper and the process involved in making it. SeedMapper is a Chrome browser extension that, each time you open a new tab, will pick a random entry from the Seed Vault’s database and generate satellite imagery from the location from which it was harvested. It is available for free in the Chrome Webstore.

The Svalbard Global Seed Vault stores seed samples from all over the world, some over half a century old. When seeds are harvested they are typically brought to a local genebank and some are further shipped to Svalbard for extra safekeeping. This provides the Seed Vault with an extremely diverse collection, representing crops, nations and the effort of thousands of people from around the globe.

In our initial research for this project we pointed at some of the dominating cultural narratives concerning the Seed Vault. These narratives are primarily about framing the Seed Vault as a backup scheme for the impending doom. However, we believe that there is a richer story to be told about the global collaborative undertaking to preserve food safety, nature and biodiversity on the planet. This design experiment is an effort in trying to tell this story about the Seed Vault by making it more available and visceral through everyday means.

In this experiment we try to connect the samples stored in the Svalbard Global Seed Vault with their place of origin, and to highlight the relationship between Svalbard and the wider world. With the extension installed, whenever you open a new tab in your browser, you will be showed a random sample from the Seed Vault’s database and a satellite image from the location of harvest. We combine this image with information about the crop, name of the country, when it was acquired and the distance in kilometres from the Seed Vault.

In this way, SeedMapper enables us to travel to distant, spectacular and often forgotten corners of the world, emphasising their beauty and frailness. It draws attention to how food safety and upholding a healthy biodiversity is both a global and local collaborative effort that needs to happen over decades and centuries.

Designing SeedMapper

One of the central constraints for the design brief we set ourselves was to make the seed vaults database more accessible and bring it into the world of the everyday. In the Seed Vault colouring book and shop we literally make the database physical and tangible through mundane objects such as cups and tote bags.

In the digital realm, the confines of the browser is a place where we spend much of our time. For a while we have been using Google Earth’s lovely chrome extension, which serves a beautiful image from Google Earth every time you open a new tab. When we discovered the geolocation data in Genesys, we saw the potential to utilise the same basic concept to embed the seed vault into the every day practices of using the Web.

While Google Earth’s extension offers a highly curated set of spectacular imagery from around the world, the SeedMapper will only take you to places represented in the Seed Vault’s database. These places are primarily agricultural locations in rural and often overlooked pockets of the world. They are, nonetheless, beautiful, and often spectacular, in their own way.

After using the extension for a while you will start to recognise some of the places, and get a sense of how a certain place, country, species or crop is more or less represented in the vault. There is a certain joy of discovering a new crop or a new country you haven’t seen in a while. Every now and then you also get dropped into the most remarkable of locations, in a country you barely knew existed. Try it yourself.

The making of

Data retrieval and processing

As we noted in our data exploration post, Genesys holds much richer data about the accessions than we can glean from the Seed Vault’s own database. We were immediately interested when we discovered that it contains information about the geolocation of the point of harvest. Unfortunately, Genesys does not contain all of the Seed Vault’s content (some countries and institutes are not part of Genesys) and geolocation data is not always available.

Genesys’ API is not really built for public use in third party apps. But we were able to download search results under a certain size to a processable CSV file. By checking ‘Safety duplicated in Svalbard’ and selecting small ranges in latitude, we were able to download all the entries in the seed vault with corresponding geolocation data, in about 10 batches.

We then combined the csv-files and filtered out any entries with flawed latitude or longitude data. We were then left with an array containing 327427 accession ids (disclaimer: processing error on our end might have excluded some or made duplicates here). With these ids we are able to retrieve the full data on that entry using the API endpoint[accession_id]. The fields we extract from this are the crop name, species common name, name of origin country, and time of acquisition. From this we construct the he overlaid text which is presented as a readable paragraph. This needs to take into consideration that any of the fields might be missing or invalid and adapt accordingly.

The plugin

The extension consists of a html and javascript application that, due to CORS restrictions, loads data from a Node Express application running on Heroku that will proxy the data from Genesys.

When called the Node application will fetch a random accession id from a json-file. It gets the geolocation data from the API endpoint at Genesys mentioned above. The latitude and longitude is used to fire up a Google Maps tile. This is overlaid with the crop name, what kind of species it belongs to, when and where it was acquired and a calculated distance from the current location to the Svalbard Global Seed Vault (Geolib is used for the distance calculation).

Show your support

Clapping shows how much you appreciated Jørn Knutsen’s story.