Remote Sensing and Geospatial Data Analytics
Toledo, OH
I am a geospatial professional with 3+ years of experience in solving diverse sets of problems. Currently, I work as a Research Scientist, specializing in data science applications for remote sensing. In my day today, I develop statistical models for tree species classification, use Py6S for radiation calibration or perform a literature review on GHG emissions, to name a few. I have held various geospatial data-centric positions where my responsibilities varied from physical data collection to curation to modeling.
SPATIAL NETWORKS
What parts of Manhattan are more connected than others?
Question
Do most connected parts of the city change with time of the day?
The aim of the project is to understand how connection within the city change. Are there any recurring patterns? I will use modularity to investigate clusters in the city. I will try to investigate these patterns and understand why such clusters are coming up.
Modularity Classification
Travel patterns vary through time of the day and day of the week. To account for these different patterns of travel the network was divided in these classes:
o The complete network
o Weekday Peak hour travel network
o Weekend travel network
o Networks based on 4 hour timesteps
The identified peak hours on weekdays were 7-11 am and 5-9 pm. Peak hours were identified trip density of all weekday travel. For all the network classes degree and Euclidean distance were calculated. Modularity for each class were generated and mapped. It was also observed that some parts of the city always lie in the same cluster throughout the day whereas most clusters kept changing throughout the day.
Results
To interpret recurring patterns in modularity classes I used landuse data along with major landmarks and median income of block groups. Results obtained from this are explained below:
o What is the travel pattern?
Spatially contiguous clusters explain that people do not tend to travel farther places very often. Distribution of distance over time shows that travel distance increases exponentially during peak hours and then reduces .
Since people tend to travel to farther location for work weekday peak hour have longer average trips. Standard deviation of weekend trips is also higher.
o Did modularity clusters change though the day?
It was observed that modularity classes changed throughout day. Specifically, in the first (00:00- 04:00) and second (04:00- 08:00) timestep there were no contiguous attern. Modularity classes start to emerge as the travel density increases with time. Most continuous clusters were found around peak hours. This pattern suggests people living in similar neighborhood tend to have similar travelling pattern during their working hour, but the pattern does not comply otherwise.
o Why is South Central part connected to farther away places in the city?
Modularity class 0 is the most vulnerable cluster. It changes with time. The cluster is spatially contiguous, with some parts in the farther parts of the city. This pattern is arising because the cluster comprises of major tourist areas. The farther away locations are all tourist places which explains the development of network between south central part of the city and the father away places.
o Why North Central part is the most spatially contiguous cluster?
North Central part of the city comprise of mainly commercial and high to middle income residential buildings. The daily commute from residential and commercial at peak hour rives rise to this cluster. This socioeconomic class the major user of yellow can rides hence the density of network is highest in this area, which explains why the cluster is contiguous.
o What is happening in the North East part of the city?
North East part of the city is one of the lower income group in the areas. Yellow cab rides data used in the exercise does not represent this socio-economic group hence the modularity of the area is not a good representation of what is happening in the area.
o Modules appearing only on weekends
While the general network pattern of the city remains constant through time, there were some trips that increases only on weekends. Most of these trips were short distance trips in the blocks with big grocery chains like target, and Walmart. These shorts distance trips on weekend explain the shopping pattern of people on weekends.
