Awards & Nominations
Sosqua has received the following awards and nominations. Way to go!
Food for Thought
SOSQUA : Harvesting beyond food
Summary
SOSQUA is a community-supported agriculture web-app that links family farmers and local markets, based on actual food production and demand. Harvesting time per farm is estimated by monitoring growing crop cycles using optical satellite imagery-series and deriving vegetation indices. Besides, a non-supervised classification is performed to identify croplands, which is validated by the farmers through the customized mobile application. This platform can be accessed by local markets to check for product availability and to contact suggested farmers, based on their location and harvesting time. Also, government institutes can access the portal to monitor harvesting dynamics and control prices.
How We Addressed This Challenge
SOSQUA is addressing the ‘Food for Thought’ challenge, specifically in regards to the actual disruptions in supply chains that entail changes in commodity prices, tending to be more severe for high-value commodities (perishable products), which are often produced by smallholder farmers.
Small family farmers are core suppliers of food & nutrition. Around 60 million people work as smallholder family farmers in Latino America and the Caribbean, who produce 80% of the agriculture sector, which represents almost 40% of the sector GDP. Only in Colombia, there are 9.4 million family farmers on the frontline keeping the food supply chain running during the COVID-19 crisis. However, restrictions on movement and non-perishables preference are keeping 90% of these families from selling products, which has triggered food losses on top of it.
A way to support family farmers during and in the post-COVID crisis is by connecting them to local markets. Answering where, when, and what products are about to be harvested (supply) and demanded can turn this crisis into an opportunity to change the agricultural market model towards promoting a win-win relationship between local farmers and consumers. SOSQUA answers these questions by assessing the growing cycle of crops from the space and alerting about harvesting calendars at a farm-crop type level, focusing specifically on small family farmer lands. Knowing early information about supply can benefit not only local governments but local markets and grocery stores to make better decisions in planning delivery chains, thus mitigating the impact of movement restrictions, besides powering the food market through monitoring food prices.
How We Developed This Project
Our approach was based on geographical decision support systems. Squeezing spatial data to get the most insight out of it.
Study area: 5 square Km, next to the Lake of Tota, Colombia, one of the territories mostly occupied by family farmers.
Data:
Two raster data series (December 2019 -May 30, 2020) were used to start with the exploration of the study area.
- Data series Sentinel, level 1A ( surface reflectance)
- Planet series, (surface reflectance)
One vector dataset download from the open GEOportal of the Colombian National institution in charge of land management (IGAC):
- Shapefile of family farmers lands (polygon)
Methods:
Both imagery data series were used to estimate a series of Enhanced Vegetation Index (EVI), which is similar to the Normalized Difference Vegetation Index (NDVI) and can be used to quantify vegetation greenness.
A python3 code was developed to process all the imagery datasets. (All our code can be found in our GitHub repository). The code is based in a loop In each operation is opened a dataset, calculated the EVI. Saved the EVI in a raster and finally used the Mask.tif to obtain the average of the EVI values per zone.
In order to consolidate EVI averages per for each crop, it is was necessary to create a layer specifying the farm areas. The mask was developed using the result of a classification non-supervised in QGIS (using the Semiautomatic-Classificaton Plug-in of QGIS), a predial bounds shape file of the zone and, finally, a raster of the layer "Likely for Family Agriculture" from https://sipra.upra.gov.co/.
We estimated the multitemporal change of the EVI index over covers by farmer-land, and used these changes as a percentage of growth and harvesting time per type of crop.
Access a tableau dashboard and geographic visor with this analysis.
Please access our website platform for enjoying the full version of SOSQUA!
Software:
QGIS, Semi-Supervised Classification, Raster Analysis
Tableau
Phyton
Project Demo
Welcome to SOSQUA by watching our video
https://youtu.be/h4XbUkR511w
Project Code
Data & Resources
Sentinel 1A: 6 scenes from December 2019 -May 30, 2020
Planet : 17 scenes