Estimating residence-time matrix

Extending Notes on: Horne, J. S., Garton, E. O., Krone, S. M., & Lewis, J. S. (2007): Analyzing Animal Movements Using Brownian Bridges, we now want to calculate the residence-time matrix for various polygons (AGEBs). To do this, we will use the geopandas library along with the data from Hermisillo that we have for a set of individuals. We will currently work with the data of just one individual.

1. AGEB data

The data for the AGEBs was provided by INEGI (National Institute of Statistics and Geography). There are a total of 582 AGEBs of which we might discard some that are not of interest. From the data, we particularly care about CVE_AGEB and the geometry - which is a polygon.

import geopandas as gpd

data = gpd.read_file("/home/boticelli/Documents/uta/code/residence-time/bbmm-drive/26a.shp")
data.head()

We can confirm that there are 582 AGEBs by checking the number of rows in the dataframe data.

  len(data)

Each GeoSeries has a corresponding CRS (Coordinate Reference System) that defines the system used for the projection. If a GeoDataFrame has only one column that corresponds to a GeoSeries, then the CRS of that series corresponds to the CRS of the dataframe. A dataframe can also have multiple GeoSeries, each with their own corresponding CRS. In such a case, one of the series is designated as the active geometry column and is used by default for any geometrical operations.

data.crs

Here, the data we have is in the “MEXICO_ITRF₂₀₀₈_LCC” system. We will now transform the system to a Pseudomercator system for ease of use. This can be done using the geopandas.GeoDataFrame.to_crs function. The EPSG code corresponding to Pseudomercator projection is EPSG:3857.

data.to_crs("EPSG:3857", inplace=True)

We can also visualize this data using plotting functionality in geopandas and matplotlib.

import matplotlib.pyplot as plt

data.plot(figsize=(5,5))
plt.savefig('/home/boticelli/Documents/uta/code/residence-time/agebs.png')