2016 conservation track: strategies and tips for large scale data collection and automation by...

Post on 16-Apr-2017

67 views 0 download

Transcript of 2016 conservation track: strategies and tips for large scale data collection and automation by...

The ParkServe™ Team – September 2016

Strategies and Tips for Large Scale Data Collection and AutomationThe Trust for Public Land’s ParkServeTM

5/17/2016

Presentation Agenda

Overview: The Trust for Public Land as an organization

Context: ParkScore® and ParkServe™

Strategies and Tips: Collection, Creation, Loading and Modeling

Conclusions: Best Practices, Roadblocks and Limitations

1

2

3

4

Our Mission

The Trust for Public Land creates parks and protects land for people, ensuring healthy, livable communities for generations to come.

The Trust for Public Land’s ParkScore®

• Taking ParkScore® to scale

• Focuses on park accessibility metric for all urban areas in the US

What is ParkServe™?

By May 1, 2018 ParkServe aims to map parks and park access for:• All 3,573 US Census 2010 defined Urban Areas-Urban Clusters

– 13,931 US Census 2010 places (including 12,762 Cities) 

– 2,494 of 3,143 US counties (70%)

ParkServe at 100% would summarize 10 minute walk park access for: 

– Over 80% of Americans  – Over 80% of American households earning less than $35,000 per year – Over 90% of all persons of color

What is ParkServe™?

Based on 2015 US Population Estimates—284,301,095 people live within Block Groups intersecting Urban Area-Urban Cluster areas of 318,536,439 total US population (Sources: US Census Bureau 2010 Census Urban Areas and Place boundaries, and Esri, Inc. Business Analyst 2015 Demographic Dataset).

What is ParkServe™?4 Components of Completion

1. Data Collection2. Data Creation3. Data Loading & Preliminary Modeling4. Data Verification & Final Modeling

Iterative Approach to Completion

ParkServeTM Collection

ParkServeTM Collection

ParkServeTM Data Creation

ParkServeTM Reporting

ParkServeTM Reporting

ParkServeTM Data Loading

ParkServeTM Modeling

ParkServeTM Data Verification

• Data Model Evolution– Start simple and add complexity as necessary

• Data Collection Workflow– More testing/development before scaling up– Speed/Progress vs. Efficiency

• Complex automation– Everything built from scratch– Learning process for whole team

ParkServeTM Roadblocks and Limitations

The ParkServe™ Team – September 2016

Thank you!Emmalee Dolfi

Emmalee.dolfi@tpl.org

Holly WinscottHolly.winscott@tpl.org