Wolfram Computation Meets Knowledge

Wolfram Summer School


David Torres

Science and Technology

Class of 2019


After finishing college in Lima as a Forest Engineer, David moved to Cuzco to volunteer in Machu Picchu Historical Sanctuary helping the park rangers, which led to his thesis research at the National Park Service on ecological systems mapping. He began working at Nature Services Peru (https://www.natureservicesperu.com/about-us) in 2014, designing and implementing Payments for Ecosystem Services schemes that connect communities from the Peruvian Amazon with companies in the main cities of Peru. Together these partnerships form the Nature Stewards Network (https://www.regenera.pe), working to reduce the deforestation in key landscapes of the country. While doing graduate school at UC Santa Cruz, Coastal Science & Policy program (https://csp.ucsc.edu), and by leveraging his time living in the SF Bay Area (aka Silicon Valley), he is developing a digital platform to scale the Network in the whole Amazon, adapt it to also support sustainability efforts in the Coast, and involving cities from other countries as well. David’s main interests are building tools and testing ideas in the real world.

Computational Essay: Hurricanes in the XXI century

Project: Rainforest stewardship through blockchain


This project focuses on the automation of forest loss detection for any property in the Amazon and its registry into a blockchain, which then will provide that information in real time to the participants of the network via periodic verifications. In order to get access to the network, a specific contract has to be previously defined. If the requirements of the contract are met, then the owner of the property will receive the agreed payment but if increasing deforestation is detected, there won’t be any payment.

Summary of Results

We achieved the standardization and automation of forest loss detection at the property level, from an already processed dataset of forest classification. Key definitions of the functions and one real example from Peru were developed. Following this logic, we can analyze the historic deforestation of any property in the world with forest cover on it, which could be used as a filter or ranking metric for existing climate change mitigation strategies or conservation incentive schemes.

Future Work

– Incorporate machine learning, convolutional neural networks, and satellite imagery with better temporal and spatial resolution, then connect these results with a blockchain through a classifier; – Automate the detection not only of deforestation but also of land use change (what happens afterwards); – Development of a dynamic Forest Index, as a practical proxy for ecosystem’s health; – Design, develop, and test a commercial prototype of a funding platform for planetary stewardship, if successful, a company will be created.