Google's Resources for Data Science in Earth Sciences

Google has developed numerous tools, and most of us have been using these resources such as Google search engine, gmail, map, Google pay, and Google play. In addition to these general services, Google offers spectacular services for data science in Earth sciences (i.e. climatology, meteorology, hydrology, etc.). Here, I have listed some of the great resources that Google offers (many of which are free) that are useful in data science and more specifically in geosciences (the field that I've been working on for the past 10 years).

1. Google Colab Notebooks

Google Colab notebook is an amazing free cloud computing resource that provides CPU and GPU computing. Offering 12 GB RAM and over 100 GB disk space, it is a decent tool for exploratory data analysis and some deep learning. It can also mount Google Drive, and the outputs can be easily saved and shared from there. Perfect tool for anyone interested in gaining experience with some sort of cloud computing and those who want to learn and work with deep learning. To summarize:

• CPU & GPU computing, 12 GB RAM, 100 GB disk space & mounting Google Drive
• Zero configuration required (no need to install any Python libraries!)
• Easy sharing

Learn more about Colab Notebooks here: https://colab.research.google.com/notebooks/intro.ipynb

2. Tensorflow

If you have worked with any deep learning tool in the past couple of years, you have most likely (at least) heard about Tensorflow. It is an end-to-end open source platform for machine learning. The high level user-friendly APIs such as Keras (among other flexible tools) let the user easily build and deploy robust ML models, and perform experiments on the model/results (e.g. Tensorboard). In addition, Tensorflow Playground offers a visual interface to learn the basics of machine learning and play with various configurations to see how they work.

Learn more about Tensorflow here: https://www.tensorflow.org/

3. Google Cloud | Google computing systems

Google Cloud consists of a set of physical assets (e.g. computers and hard disk drives) and virtual resources (e.g. virtual machines) that are contained in Google's data centers around the globe. It offers very many different products such as cloud storage, compute engine, AI and machine learning, BigQuerry, Cloud SQL, and many more. Google offers a 90-day free trial (up to $300 cloud and computation cost) to anyone who is interested.

A particular advancement that cloud computing in general (and Google cloud in particular) has provided in the past few years is the possibility to access vast resources at a relatively low cost. For instance, when I was doing my PhD in 2013 and working on climate change impact assessment, I had to download terabytes of data on the Linux systems in our university (that costs tens of thousands of dollars and requires continuous maintenance), and then analyze the data with those computing resources. Due to excessive cost and maintenance demand, performing such analyses was not feasible for small businesses or consulting firms. Cloud computing has completely changed this pattern and allowed businesses to gain a huge competitive advantage against conventional academic approach. Notably, the most-recent set of climate model projections (i.e. CMIP6) is now freely accessible on Google cloud, and a few startups have already analyzed, downscaled, and utilized the data for climate change impact assessment (e.g. Descartes Labs or Jupiter Intelligence). The type of study that would have taken weeks to months to carry out in a university setting now takes minutes to hours on Google Cloud. 

Learn more about Google Cloud here: https://cloud.google.com/products/

4. Google Earth Engine (GEE)

GEE is a computing platform that allows users to run geospatial analysis on Google's infrastructure. It offers a web-based IDE code editor for writing and running codes, and users can leverage the Earth Engine JavaScript API. GEE offers efficient access to several satellite observations including Landsat, Sentinel, and MODIS, as well as various hydro-climatological datasets (reanalysis) such as NLDAS, GLDAS, TerraClimate, ERA5, NEX-GDDP, and more. Freely available for research, education, and nonprofit applications, it is an invaluable tool to advance science.

Learn more about Google Earth Engine here: https://earthengine.google.com/

5. Google Dataset Search 

Simply search for any data and find relevant links for it, whether it is precipitation, population, or even plane crashes.

Link to Google dataset search: https://datasetsearch.research.google.com/

6. Machine Learning Crash Course

Featuring a series of lessons with video lectures, real-world case studies, and hands-on practice exercises, the crash courses provide a self-study guide for aspiring machine learning practitioners. It provides a review for ML concepts, ML engineering, and ML systems in real-world systems (e.g. cancer prediction). 

Learn more about crash course here: https://developers.google.com/machine-learning/crash-course

7. Google Tech Dev Guide

Google’s Guide to Technical Development is a great resource for those interested to be hired at Google as a computer scientist or software engineer. It provides a collection of materials from many sources and coding questions that are asked in Google job interviews on foundations of programming, advanced programming, machine learning, and cloud computing. 

Learn more about Tech Dev Guide here: https://techdevguide.withgoogle.com/

8. Google Career Certificates

Google is setting a new standard for degree requirements. It offers certificates that take only 6 months and significantly cheaper than conventional university degrees (only $49 per month). The certificates are to be treated as a bachelor's degree for Google's own hiring. The subjects covered are as follows:

• Data analyst
• Project management
• Web (UX) designer
• IT specialist

Learn more about Google career certificates here: https://grow.google/certificates/