Exporting GIF of Time-Series Trend Analysis from Kriging

From CUOSGwiki
Jump to navigationJump to search

Introduction

Purpose

Animating data has the ability to engage viewers in ways numerical data cannot, and can demonstrate how a phenomenon is connected and evolves through a series of events. The ability to visualize meteorological data via a time series animation, provides viewers with an option to see the progression of cold fronts, heat waves and precipitation, and can assist in forecasting events allowing agriculture specialist to plan crop planting, irrigation, harvesting and chemical application. In the example that will be demonstrated in this tutorial, precipitation data will be used to create a time series event that would be useful in reviewing historical data that can be used in cross analyzing weather events with floods or droughts, and to detect patterns.

Jupyter Lab will be the platform used to create the time series animation, in conjunction with Anaconda, Pandas, and PyKrige. Using Jupyter Lab will enable the user to develop skills in scripting, and is a more efficient way to complete our time series animation. Jupyter is a flexible open-source environment for notebooks, code and data that is capable of handling a wide variety of file formats. Anaconda will provide the building environment to enhance ease of collaborative work and to ensure a homogeneous work environment. Anaconda is preferred for our tutorial due its ease of use for beginner programmers with its pre-installed packages that that are helpful for data science and machine learning. Pandas is a package built on top of another package (Numpy), and will facilitates data analysis by providing python functions and methods (specifically for Time Series creation); and lastly PyKrige provides a Python library that is used for geostatistical tasks such as interpolation, which we will use to create new points to provide a smoother time series animation.

The intended audience of this tutorial are those that existing knowledge and experience using Python or Anaconda, Notebooks (CoLab or Jupyter) and Pandas.