Skip to main content

· 9 min reading
Alejandro Coca-Castro

alt=Timeline

Hello world

In the previous blog post I described my trajectory from undergraduate to PhD, which was undoubtedly a satisfactory stage with a lot of personal growth and countless academic and professional achievements. For the second entry I will share what I have experienced during the pandemic, a very difficult period for all of humanity, but one that reminds us how fragile we are like any other living species on earth.

Let's continue!

Stage 3 Pandemic and uncertainties

Let's start in March 2020, when COVID-19 restrictions started in most countries and coincided with the deadline for submitting my PhD thesis. Moreover, it was the month in which my sponsor's grant funds came to an end. Not having many options, my wife and I decided to leave London and return to Bogota (my hometown). Despite all possible pressures, I fulfilled the purpose of submitting my thesis on March 30, 2020, a very memorable day! Subsequently, in May 2020 I defended my thesis remotely celebrating by video call with my family, supervisors and friends. I then received my PhD degree in Geography from King's College London in September 2020.

During the period of my thesis corrections and job search I was selected in a virtual training program in data science sponsored by the Colombian government and organized by Correlation One. As part of the program, with other participants we applied what we learned in a project focused on the development of models and a data dashboard to predict recidivism in the Colombian penal system (see details here ). After, in August 2020, I became a contractor for the Instituto Geográfico Agustín Codazzi (IGAC), a Colombian government entity known for its role in cartography and land use planning. In my 5-month contract I had the opportunity to be the technical leader of groups of 5 to 6 researchers and to publish in scientific journals results of two projects related to data science and artificial intelligence technologies for building mapping in rural environments (see details here) and digital soil mapping (see details here). I also coordinated a Correlation One data science challenge that aimed to develop predictive models from a soil database curated by IGAC (see repository here).

With the end of my contract at IGAC and not having much success in the search for options in academia, I decided to take the path of entrepreneurship. On this path, it is worth mentioning that the learning I received in an extracurricular program offered by the entrepreneurship institute of my university during my PhD was valuable. Thus, I registered GeoMagic Labs, a geolocation technology consulting company, within the Bogota Chamber of Commerce. With GeoMagic Labs I was selected in a consulting competition with Solidaridad Network, an NGO focused on facilitating the development of socially responsible, environmentally sound and profitable supply chains. In my proposal I developed a methodology for mapping coffee production systems using open access satellite images (see details here).

Opportunity 4 - Data Science Training
Data science has undoubtedly gained an important role in the different domains of knowledge with impact in all sectors. Because of this, a number of training programs and resources are available. In my case, taking advantage of the period of uncertainty and virtuality, I started looking for ways to improve my data science skills. In this way I found the Data Science for All (DS4A) programme from CorrelationOne. Sponsored by the Colombian government, this program offers 13 weeks of training open to all Colombian citizens who are interested in strengthening their data science skills and joining the growing artificial intelligence ecosystem in Colombia. To participate in the program, students must pass an admission test that evaluates general knowledge of statistics, programming and logical reasoning. The program is very interesting, as not only do you receive training, but you also have talks with professionals from industry and academia telling how they use data science in their day-to-day work. I can describe the program as basic to intermediate level with very practical examples and the possibility of working in teams to put all the knowledge into practice through an applied project proposed by government institutions, private sector or NGOs. CorrelationOne offers other training programs focused on women, empowerment, engineering, among others that are open to any nationality.

Stage 4 Open Science, May 2021 - to present

In November 2020 my PhD sponsor's condonation rules changed opening doors to new opportunities. It was no longer mandatory to prove that I was linked to an institution in Bogota for the time I received funding (that is, 4 years!). The new regulations were flexible with other types of deliverables including publishing articles and/or starting a small and mid-size enterprises related to my PhD topic. Under this new scenario, I applied for a postdoctoral position at the Alan Turing Institute in the UK in December 2020. I was already familiar with the Turing environment from a previous collaborative hackathon opportunity I had participated in December 2019 (see the report here). The postdoctoral position fit my profile, and I was attracted to the possibility of learning probabilistic modeling, improving skills in software development, and contributing to Pangeo, a community that promotes open science for big data analysis in geoscience.

In February 2021 I received a communication for an interview which included a 5 min presentation, followed by an explanation of code of my authorship and questions from the interview panel. It was my first postdoc interview so it was natural to be a little nervous. Nevertheless, I offered accurate answers highlighting my experience collaborating in various research groups. At the time of showing my code I managed to demonstrate how it had been used or visited in Zenodo (a platform to register any digital research object). A few weeks later I received the offer of the position starting remotely from May 2021. Later, with my wife, we left our life in Bogota and decided to restart our life in London at the end of July of the same year.

Being part of Turing, I have had the opportunity to collaborate with multiple researchers in the Uk and their networks, below are the projects or initiatives that I actively collaborate with:

  • My postdoctoral project involves the development of tools based on intelligent fusion of environmental sensor data to improve the monitoring of our changing planet (see details here). In collaboration with researchers from the British Antarctic Survey and Met Office we are developing a tool based on probabilistic modeling that combines data from different modalities (grid data, in situ and auxiliary stations) to have finer and more local-aware observations of environmental variables. In addition, with greedy algorithms, the tool will suggest where to install local sensors to reduce the uncertainty of the modeled environmental variable. I am fortunate to be able to contribute with my expertise in the development of the tool, in my case working with researchers at the UK Centre for Ecology and Hydrology (UKCEH) to model soil moisture.
  • Co-founder and leader of the Environmental Data Science book (see details here). It is a resource led by the environmental community where Jupyter notebooks are used to demonstrate advances in data, methodologies and other types of research using open source tools. Through this initiative I collaborate with other members of the Pangeo community in Europe in order to increase the democratization of scalable and open source tools for geoscience.
  • Member and developer at scivision (see details here), an open source tool developed among researchers from different domains, including molecular biology, agriculture and environment, as well as professionals from the Turing Research engineering team. My role in the project is to ensure that the tool is generalizable to a variety of imagery (satellite, drones, climate data, among others) and models from the environmental science community.
  • Member and co-leader of the translation team at The Turing Way, a community-led initiative to promote best practices in open and reproducible science (see details here).
Opportunity 5 - Open Science Communities and Training
Beyond improving my skills and best practices in programming for environmental applications, the last stage has allowed me to be part of and contribute to the open science community. Despite having participated in multiple research groups since my undergraduate years, I think that open science communities have the greatest potential to have a long-term impact on our society. Examples of these communities are The Turing Way and Pangeo, the first focused on a general audience and the other on the geoscience community. They are communities that promote a vision of science in which anyone can contribute to a common goal, democratizing knowledge and good practices in projects in different productive sectors. The Turing Way has a diverse communications channel (see details here), but what I advise in the first instance is to join through Slack. With nearly 400 members, this is the place where you can exchange ideas with other members and learn about community activities. During the course of the year, `book dash` weeks are organized, which are hackathon-type events in which various activities are proposed, such as proposing or reviewing online resource content, short talks on various open science topics and socializing or networking spaces among attendes. Unlike traditional hackathons, there is no competition for any prize, any contribution is part of a common celebration. Through The Turing Way, I also learned about the program Open Life Science (OLS). It is another community which offers 16 weeks training in open science leadership. I participated in the fourth cohort to validate the idea of the environmental data science book through meetings with my mentor. Later, I co-mentored a multi-language open science knowledge project in Central Asia in the fifth cohort. Both experiences were valuable, since in the first one I was able to learn the basics in leading an open science project, and later I returned what I learned through my mentoring role. I fully recommend joining OLS, they are open to any knowledge domain in which you want to practice open science.

Summary

We have completed the second entry of Alejandro 101. I am pleased to share my professional and academic development, as well as some opportunities for those interested in open environmental science and information technology.

If my content is relevant I will be publishing monthly various topics from discussions of tools (with code) as well as concepts or schemes proposed in open science, environmental or information technology topics. I will also share relevant updates of my personal life or travels that may be enriching for other readers of this blog.

For an inclusive and transparent science 🚀

· 8 min reading
Alejandro Coca-Castro

alt=Timeline

Hello world

Motivated by the timeline with which I participated in the fireside chat organized by the Turing Way the first entry of my blog is a reflection of what has been my journey in the last 11 years.

I will summarize my most relevant experiences in open environmental science and information technologies through a series of anecdotes. These anecdotes go beyond the technical aspect and include opportunities that I share for others identified with similar paths to take advantage of.

Let's get started!

Stage 0 beginnings, < 2011

Although I did not reflect it in my timeline, since my undergraduate studies in Agricultural Engineering I got my first interactions in spatial information technologies. I had the opportunity to study subjects related to spatial data and models as part of the Geomatics postgraduate program at my alma mater, the National University of Colombia. I also did an exchange with the University of São Paulo, Brazil, where I learnt a new culture and participated in research groups using geoprocessing technologies applied to environmental and agricultural studies. In this exchange I was trained in several satellite image processing tools, both closed license tools such as ERDAS and ENVI, and open license tools such as SPRING. Also, my GIS skills were mainly in ArcGIS, and a little in QGIS. Ya finalizando mis estudios, en el mes de Mayo de 2011 tuve la oportunidad de poner a prueba mis conocimientos con mi primera consultoría realizando una base de datos y cartografía para un catálogo de plantas medicinales en Colombia (ver detalles aquí).

Opportunity 1 - International mobility
Do you know that most undergraduate programs have exchange programs with international universities? For example, the Office of Interinstitutional Relations (ORI) at the National University of Colombia has a series of agreements with multiple universities worldwide. Thanks to these agreements, I was able to spend 6 months in Brazil, and also received support to participate in a conference at the XV Simposio de Sensoriamento Remoto de Brasil, in which I presented research results from one of the groups I was involved with (see the article here). If you are an undergraduate or graduate student, I invite you to take a look at these programs that often go unnoticed, but are very beneficial to learn about cultures and ways to apply science in other contexts.

Stage 1 open deforestation data, 2011-2014

As part of my undergraduate internship I joined the Terra-i project at the International Center for Tropical Agriculture (CIAT), now known as the Alliance of Bioversity-CIAT. This project had a very positive impact on my professional development. Terra-i is an initiative for the global monitoring of habitat loss by generating free and open deforestation alerts. Debido a la relevancia del proyecto con la agenda de la organización, participé en varios subproyectos que usaban los datos de la herramienta para cuantificar el estado de los ecosistemas en Latinoamérica y Caribe así como para pronosticar el impacto de proyectos de infraestructura vial en la región (ver detalles aquí).

As a proactive researcher I represented the project in several international events on conservation, forestry and geospatial technologies. I also extensively promoted the use of open data through case studies on the project blog (see for example ). I was the author and coordinator of a protocol for field validation of deforestation alerts in the Peruvian Amazon, which was later replicated in other regions. In terms of communities, I was a scholar of the Society for Conservation GIS (SCGIS), a program funded by ESRI and environmental organisations, which I personally recommend to improve knowledge in GIS tools and meet other researchers worldwide applying these technologies for conservation.

Opportunity 2 - GIS training programs
As I mentioned before, I had the opportunity to be a scholar of the SCGIS program. The application is open annually and selects professionals working in the conservation area to improve their knowledge in GIS tools, mainly ArcGIS (see details here). In addition, the program allows grantees to present their work at the ESRI user conference as well as the annual SCGIS conference. The process is very competitive but with good motivation, preparation and evidence of conservation work on the application form it is possible to be selected. It is important to point out that the program partially covers the costs, and therefore it is necessary to count on external support for certain expenses such as airline tickets and visa if needed. Other training programs or summer schools can be found on different topics in addition to conservation. My advice is to look for those where you have a greater affinity, and where they offer the possibility to stay connected with instructors and colleagues who are part of your cohort.

Stage 2 MSc. and PhD in Geography, 2014-2020

As part of the Terra-i project, I met Mark Mulligan, professor at King's College London with whom in collaboration with the Terra-i project we studied the causes of deforestation in the Amazon biome. It was a period where I trained and applied data science, artificial intelligence and cloud computing. De esta manera, en mi maestría analicé los patrones espaciales de la deforestación mediante el uso de minería de datos y análisis fractal (ver detalles aquí). From the use of two open data sources of deforestation, Terra-i and the Global Forest Change maps, my research was able to map the spatial distribution of spatial patterns associated with different agents of deforestation. En el doctorado tuve un abordaje más profundo con la aplicación de técnicas de análisis de las series de tiempo para extraer información de las coberturas y uso posterior a la deforestación a partir de imágenes satelitales (ver detalles aquí). In this research we innovated in the use of convolutional recurrent neural networks for time series classification using MODIS images. The models were trained using hardware resources from the Terra-i project as well as Google cloud computing credits.

Beyond the technical advances during my studies, I think the most exciting experience was collaboration and mobility (exchanges, internships, conferences and research groups). Por ejemplo, en mi maestría hice un intercambio en Global Canopy, una ONG ambiental con sede en Oxford, Reino Unido en la cual hice un análisis de la seguridad hídrica en el Amazonas (ver detalles aquí). In my PhD, I did exchanges in 2011 at HEIG-VD in Switzerland and later in 2019 at CIAT's Asia regional offices in Vietnam. At the first institution I received training in Python programming, Big Data and Deep Learning in the lab led by Professor Andres Perez-Uribe, and at the second I refined the analysis and modeling with my co-supervisor at CIAT, Dr Louis Reymondin. I would also like to highlight the summer 2018 internship held in Satellite Application Catapult at Harwell Science and Innovation Campus (UK) where I participated in the accelerator Frontier Development Lab (FDL). En esta aceleradora participé con otros investigadores afiliados en instituciones de Europa para el mapeo de asentamientos ilegales a partir de imágenes de resolución media como Sentinel 2 y muy alta como WorldView-4 (ver detalles aquí).

Opportunity 3 - Free research credits for cloud computing from Google and Microsoft
In addition to the hardware resources provided the Terra-i project, I had the initiative to explore cloud computing through the education/research programmes offered by Google and Microsoft. Google has a research credit program. On the other hand, Microsoft has the AI for Earth program which offers Azure platform credits for environment-related research. Unlike Google's generalized program, the AI for Earth go beyond providing the cloud service and aim to connect all researchers and initiatives by proposing methodologies based on data science and AI to make an impact on their environmental projects. I am not in favor of in a particular program as each has its advantages according to the needs of the research and preferred community. For both programs, you are required to complete an application form that asks a series of questions of the research and an estimate of the total credits required. To maximize the chances of being selected, my suggestion is to take advantage of the trial subscription of each cloud provider. In this way, you can make a better estimate of the credits required as well as additional services to be mentioned in the application.

Summary

In this first installment of Alejandro 101, I summarized my experiences from my undergraduate to doctoral studies. There were different people who inspired me and I was able to collaborate during all these years. It was a period full of many professional and personal achievements with some very valuable friendships that I still keep, as well as the satisfaction of being able to start the Coca Calderón family during the third year of my PhD.

In the next instalment I will continue with a second part of the timeline, from which I believe I am successfully building new foundations for leadership in open environmental science. I will also discuss opportunities and experiences that may be of interest!

For an inclusive and transparent science 🚀