[CSCI 3665] Interactive Data Visualization

Project Portfolio

Project Visualization

Data Specifications

1. Introduction

   The data we are working with is a measurement of different agricultural inputs, such as types of crops, temperatures, locations, etc. across the globe from 1975 to mid 2000s. The collaborator of this project is Professor Erik Nelson, from the Bowdoin Department of Economics and Professor Clare Bates Congdon, from the Bowdoin Department of Computer Science. Their paper can be accessed here. This project is also a part of the Agricultural and Resource Economics Commons. One of the focuses of the data is to explore the possible effects of global warming on crop yield globally and regionally. This is one of the major concerns that many scientists have been studying for many years. The dataset is very useful for addressing those concerns in terms of crop yield because it provides rich information that can be used both longitudinally (over time) and cross-sectionally (at an instant time). Furthermore, this project is important because as global warming increases, it will be particularly important to understand the major factors that affect crop yields. By understanding these factors, one can make informed decisions to sustain crop yield growth admist the rise of global warming.

   In this project, we strive to provide an interactive data visualization for our collaborators to understand more about the data, especially things that might be missed by econometrics analyses. Nelson is particularly interested in the change of crop yield over time across the globe, and would like to know what factor(s) may explain the trend. Given his goals in mind, we have come up with a list of potential specific questions regarding this "change" (see below). The research by the collaborators has already produced some findings with static images. Our goal is thus trying to come up with ways to combine these images into an interactive visualization to present a comprehensive understanding of the project.

2. Data Description
   
   The attributes that we are going to work with can be categorized into following:

   • Measurement of crop yields
   • Growing season weather
   • Crop choice
   • Investment in irrigation capability, land and machinery
   • Agricultural technology
   • Fertilizer use
   • Crop soil quality

   Each category contains more specific attributes that is described in more detailed here. The two most important aspects of the data, as we mentioned above, are the time and country components. Additionally, some of the other inputs will be kept constant so as to allow the users to investigate the effect of other inputs, ceteris paribus. The more inputs we allow to change, the more complications it will add to the understanding of each input's effects on crop yields.
   
   
3. Initial brainstorming for potential questions
   1. How did crop yield change across the globe from 1975 to mid 2000s?
   2. How did the change in temperature (day and night averages) affect crop yield over time globally and regionally?
   3. What types of crop are more popular in temperate and tropical regions?
   4. Did regional use of fertilizer have a positive effect in amount of harvest over time? If so, how much?
   5. How did specific crop types change according to the change in temperature (day and night averages)?
   6. How did specific crop types change over time globally? Is there a shift of focus on what should be produced?
   7. Did investment in agricultural technology help with growth? (yield/investment)?
   8. What agricultural input has the strongest effect on crop yield? Does this change over time?
   9. Did the size of arable land change over time? If so, was it related to irrigation?
   10. Is soil score a good indicator of productivity/fertility of the land? We could look into crop yield, along with fertilizer use and degree of irrigation, and how they could counter bad soil score among countries.

   Among these questions, there are three questions that we are particularly interested in, and would like to pursue (listed in order of importance):

   1. How did the change in temperature (day and night averages) affect crop yield over time globally and regionally?
   2. How did crop yield level of a country change as related to global crop yield (million kcal/ha)? How did this relationship change over time?
   3. How did specific crop types change over time globally? Is there a shift of focus on what should be produced?

   We will first try to implement visualizations for the first two questions; then, if time permits, we will move on to implement visualziation for the last question.

External Links
To Son's homepage
To James' homepage
To Do's homepage