# THIS CONTENT IS MOVING! How To Design and Develop a Workflow For NDVI Over Time

We are moving our course lessons to an improved textbook series. All of the same content will be improved and available by the end of Spring 2020. While these pages will automagically redirect, you can also visit the links below to check out our new content! Our course landing pages with associated readings and assignments will stay here so you can continue to follow along with our courses!

# How To Design and Develop a Workflow For NDVI Over Time - Earth analytics python course module

Welcome to the first lesson in the How To Design and Develop a Workflow For NDVI Over Time module. Learn how to design and develop automated workflows to calculate NDVI time series in Python.

## Learning Objectives

After completing this tutorial, you will be able to:

• Explain how you can use a time series of normalized difference vegetation index (ndvi) to investigate ecological processes and changes.
• List the key steps in designing data workflows.

## What You Need

You will need a computer with internet access to complete this lesson and the data for week 10 of the course.

Or use earthpy et.data.get_data('ndvi-automation')

## Normalized Difference Vegetation Index (NDVI) Over Time

This semester you have have used vegetation indices to study changes due to a disturbance such as a flood or a fire. You used NDVI to identify where vegetation increased or decreased after the disturbance event.

NDVI can also be used to understand seasonality - when an area begins to “green-up” or grow after the winter cold period and “brown down” when vegetation in an area begins to die back or senesce in an area (usually in the fall and winter).

Different areas in different parts of the world have different seasonal patterns. For example, the NEON SJER (San Joaquin Experimental Range) in California has an early green-up date on average, a short growing season due to hot temperatures and lack of precipitation and an early brown down. Whereas Harvard Forest in Massachusetts has a longer growing season and a later green-up period.

## Changes in Seasonality And Changes in Climate

Changes in seasonality can be important indicators of ecological change. For example, if green-up begins earlier, such that fruit or seed resources become available sooner than average, animals that forage on the fruits in the spring must either migrate earlier to use these resources, or miss out on them if they do not adjust migration behavior. This phenomenon is referred to as a phenological mismatch.

Designing workflows to process and create outputs for many large files and datasets is a key skill in Earth science. Yet it’s something that most scientists and earth analysts learn on the fly at some point in their careers. Here, we outline several steps associated with designing a workflow that can help you structure your thinking and develop an effective design.

### Identify your problem, challenge or question

To begin you need to clarify the question(s) or challenge(s) that you need to address. Knowing the specific problem that you need to solve will help to set bounds on what your workflow does and does not need to do.

### Identify the data needed to address that question

Once you have a question in mind, it’s time to figure out what your data requirements are. Requirements are the qualities that your data need to have in order to address a particular question well. In this week’s class your goal is to create a plot and output csv files of NDVI for your study area for a year. You want to explore seasonal patterns of “green-up” and “brown down” and thus you need data that is at least collected each month.

What data do you select for your analysis? Consider temporal frequency of data collection and spatial resolution. For example, NAIP is collected every other year, so it is unlikely to provide good information on seasonality.

In contrast, MODIS and Landsat may be more useful, because they are collected on a daily and bi monthly frequency respectively. In terms of resolution, MODIS pixels could be too large depending upon your study site. Perhaps Landsat has an ideal combination of temporal frequency, resolution and spatial coverage.