CS 594 / CS 690 - Assignment 10

CS 594 / CS 690 - Assignment 10

For this assignment, you must work in groups of one or two students. Each person is responsible to write their own code, but the group will (together) discuss their solution. In this notebook, we provide you with basic functions for completing the assignment. Complete the assignment in this notebook. You will need to modify existing code and write new code to find a solution. Each member of the group must upload their own work (i.e., a notebook file) to GitHub.

Note: Running a cell will not rerun previous cells. If you edit code in previous cells, you must rerun those cells. If you are having trouble with undefined errors and code changes not applying, we recommend using Run All to avoid any errors results from not rerunning previous cells. You can find this in the menu above: Cell -> Run All

This week, we are looking at an example of processing and analysing data in the domain of soil science. Soil moisture is a critical variable that links climate dynamics with water and food security. It regulates land-atmosphere interactions (e.g., via evapotranspiration - the loss of water from evaporation and plant transpiration to the atmosphere), and it is directly linked with plant productivity and survival. Information on soil moisture is important to design appropriate irrigation strategies to increase crop yield, and long-term soil moisture coupled with climate information provides insights into trends and potential agricultural thresholds and risks. Thus, information on soil moisture is a key factor to inform and enable precision agriculture.

The current availability in soil moisture data over large areas comes from remote sensing (i.e., satellites with radar sensors) which provide daily, nearly global coverage of soil moisture. However, satellite soil moisture datasets have a major shortcoming in that they are limited to coarse spatial resolution (generally no finer than 25 km).

There do exist at higher resolution other geographic datasets (e.g., climatic, geological, and topographic) that are intimately related to soil moisture values. With a group of soil scientists at the University of Delaware, we have paired soil moisture data and various topographic parameters in order to downscale (i.e., increase resolution) the originial soil moisture product.

This assignment walks through some early stages of data downscaling as a means to introduce you to a common and powerful Python package for data analysis: Pandas. The workflow consists of stages--data processing, modeling to generate fine-scale predictions, and visualization--wrapped in a Jupyter Notebook to demonstrate the sharability and reproducibility we discussed previously.

For the initial part of this assignment, we will process data in Pandas DataFrames to prepare it for modeling. We are looking at data for the Mid Atlantic Coastal Plains, a North American ecoregion containing the state of Delaware (designated by the Commission for Environmental Cooperation).