2e:null 2f:["$","title",null,{"children":"🌳 Tree inequity: The wealthier and whiter a neighborhood is, the greener it gets | Vy | Observable"}] 34:T486,Trees are an irreplaceable part of our planet. In addition to mitigating climate change, trees also help with bettering our lives and well-being. Nevertheless, in a world that is warming rapidly, having access to greenery and fresh air no longer seems like a basic right, but more so a battle of who has such a privilege. In this notebook, I want to explore how equitable the state of Rhode Island is in terms of the distribution of trees and greenery in its neighborhoods, using the data compiled by the [Tree Equity Score](https://www.treeequityscore.org/). **Tree equity** is defined by the Tree Equity Score organization as "an indicator of whether an area has a sufficient amount of tree canopy cover distributed in a way that all residents can experience the climate, health and other benefits that trees provide." By compiling the tree coverage percentage and socio-economic characteristics such as household income, racial diversity, unemployment, overall health, etc., we can draw meaningful insights on whether social-economic factors play a part in determining the tree equity of communities, especially of those most marginalized and vulnerable.35:T769,## Possible Insights/Hypotheses * **Possible insight 1:** Rhode Island neighborhoods with higher median household income and are predominantly white tend to enjoy more tree canopy coverage than neighborhoods with lower income and more percentage of people of color. * **Considerations:** * To test this I'll need the income (quant) and tree canopy % (quant) data for each RI neighborhood. I can then calculate the possible correlation between income/race and tree coverage. I may also want to group the income into different income groups (ordinal) to make it easy to look at, then calculate the average tree canopy % per income group. * In terms of idioms, I would need a way to visualize the possible correlation between income/race and tree coverage. Scatterplots with a linear regression and correlation value would be ideal to achieve this. To better identify which specific income group/percent of people of color enjoy more tree coverage, I can also use bar charts to display the average tree canopy % per category. * **Possible insight 2:** Residents in Rhode Island neighborhoods with less tree canopy coverage tend to experience more prevalent mental, physical, respiratory, and cardiac health challenges compared with those who live in greener neighborhoods. * **Considerations:** * To test this I'll probably need a total % of reported health challenges (quant), as well as the tree canopy % (quant) for each RI neighborhood. Similarly, I can also calculate the possible correlation between the percent of health challenges vs. tree coverage. * For the idiom, I can also use a scatterplot with linear regression to illustrate this correlation. * In this notebook, I will be testing out **insight 1** to explore whether the level of income/percent of people of color of a neighborhood correlates with the percent of tree coverage in the state of Rhode Island.36:T458,## Abstraction ### Users * Residents and officials of Rhode Island who want to assess the tree equity of their community and to improve it * Anyone who is curious about how trees and greenery play a part in leveraging the equity of urban neighborhoods ### Data Raw data was downloaded on Feb 11, 2022. * Data source: https://www.treeequityscore.org/data/ (Rhode Island) * Dataset type: Table/Spatial * Data availability: Static **Attributes used:** * **medhhinc**: the median household income of the block group * Type: Ordered - Quantitative * Orientation: Sequential * **pctpoc**: the percent of people of color inside the block group * Type: Ordered - Quantitative * Orientation: Sequential * **tes**: the tree equity score of the block group * Type: Ordered - Quantitative * Orientation: Sequential Full attribute descriptions can be accessed by downloading the data from the source. **Derived attributes:** For the bar charts, the *median household income* and *people of color %* were aggregated into pre-defined groups (Ordered - Ordinal). No other attributes were needed to be derived.37:T461,{ const base = vl .markCircle({ tooltip: true, size: 80 }) .encode( vl.x().fieldQ("treecanopy").axis({ format: "%" }).title("Tree Canopy %"), vl.y().fieldQ("medhhinc").axis({ format: "$s" }).title("Median Household Income"), vl.color({ value: "#9aa071" }) // vl.color().fieldQ("tes").title("Tree Equity Score").scale({scheme: 'redyellowgreen'}), ) .title({ text: 'As median household income increases, so does tree canopy coverage', anchor: "start" }) const line = vl .markLine({ color: 'firebrick' }) .encode( vl.x().fieldQ("treecanopy"), vl.y().fieldQ("medhhinc") ) .transform( vl.regression("medhhinc").on("treecanopy"), ) const text = vl .markText({ color: 'firebrick', x: "width", y: -10, align: "right" }) .transform( vl.regression("medhhinc").on("treecanopy").params(true), vl.calculate("'R²: '+format(datum.rSquared, '.2f')").as("R2") ) .encode( vl.text().fieldN("R2") ) return vl .layer(base, line, text) .width(800) .height(400) .data(raw_data) .render(); }38:T4f4,#### Insights * Rhode Island neighborhoods with higher median household income tend to benefit from more tree canopy than neighborhoods with lower income. * Looking at the scatterplot, there is a **low positive correlation** (R-squared value of 0.21) between median household income and the tree canopy percentage, suggesting that income somewhat plays a part in how much greenery a neighborhood gets. * In the bar chart, we can see that neighborhoods with more than $100k median income have almost 48% tree coverage - at least 21% more than those with income below the $45k ballpark. * These findings are consistent with the [national](https://www.nytimes.com/interactive/2021/06/30/opinion/environmental-inequity-trees-critical-infrastructure.html) landscape of tree inequity, with major urban cities across the country sharing the same pattern of green segregation. * In urban cities across the U.S, redlining and income segregation have for decades been disadvantaging lower-income residents. The gradient from urban to suburban could also play a role, since richer households have fled to the suburbs with greener and larger plots, leaving lower-income communities stuck at city centers and other undeveloped neighborhoods that have more pavement and fewer trees.39:T468,{ const base = vl .markCircle({ tooltip: true, size: 80 }) .encode( vl.x().fieldQ("treecanopy").axis({ format: "%" }).title("Tree Canopy %"), vl.y().fieldQ("pctpoc").axis({ format: "%" }).title("People of Color %"), vl.color({ value: "#9aa071" }) // vl.color().fieldQ("tes").title("Tree Equity Score").scale({scheme: 'redyellowgreen'}) ) .title({ text: 'The more racially diverse the neighborhood is, the less tree canopy coverage it gets', anchor: "start" }) const line = vl .markLine({ color: 'firebrick' }) .encode( vl.x().fieldQ("treecanopy"), vl.y().fieldQ("pctpoc") ) .transform( vl.regression("pctpoc").on("treecanopy") ) const text = vl .markText({ color: 'firebrick', x: "width", y: -10, align: "right" }) .transform( vl.regression("pctpoc").on("treecanopy").params(true), vl.calculate("'R²: '+format(datum.rSquared, '.2f')").as("R2") ) .encode( vl.text().fieldN("R2") ) return vl .layer(base, line, text) .data(raw_data) .width(800) .height(400) .render(); }3a:T56b,#### Insights * Rhode Island neighborhoods with higher percentages of people of color tend to receive an inadequate amount of tree coverage compared to predominantly white neighborhoods. * Looking at the scatterplot, there is a **low positive correlation** (R-squared value of 0.26) between the percent of people of color and tree canopy percentage, suggesting that there is indeed a relationship between racial diversity and the amount of greenery a neighborhood gets invested in. * In the bar chart, we can see that trees cover almost half of the surface in predominantly white neighborhoods, compared to around one-fourth in communities of color with more than 50% POC. * Similarly to policies that disfavor lower-income populations, discriminatory practices against people of color explain the lack of investment in developing the infrastructure and greenery in communities of color. For example, according to the [New York Times](https://www.nytimes.com/interactive/2021/06/30/opinion/environmental-inequity-trees-critical-infrastructure.html), up until the 70s, unfair lending decisions, health care, and infrastructure investments were made based on where people lived, a practice of redlining. The literal "red lines" that were drawn around predominantly Black and immigrant neighborhoods often line up very closely with maps showing a lack of tree canopy today in New York City.3b:[{"id":"49e70d2248fd138177b40038c83ca8020e2823826b063b831c8ec46a68beae509cde939bf97f0bc7cf05b430512c8ee5a89a02cf2a9c8342718af1b32c673ce0","href":"https://static.observableusercontent.com/files/49e70d2248fd138177b40038c83ca8020e2823826b063b831c8ec46a68beae509cde939bf97f0bc7cf05b430512c8ee5a89a02cf2a9c8342718af1b32c673ce0","name":"ri@1.csv","type":"text/csv","lastModified":1645136831730.927,"size":248685}] 32:{"id":"9d4f6c670cbdd668","slug":"tree-inequity-in-rhode-island","title":"🌳 Tree inequity: The wealthier and whiter a neighborhood is, the greener it gets","thumbnail":"94503f850e4385fe7f6e965dcec4d89e9f06e0d0e9f4fb2ae7edf296ab93a9aa","access_level":"public","create_time":"$D2022-02-11T22:04:37.582Z","update_time":"$D2024-12-23T12:44:47.233Z","view_time":null,"first_public_time":"$D2022-02-21T04:49:30.225Z","has_unpublished_edits":null,"fork":false,"stars":3,"starred":false,"owner":{"type":"team","id":"70a28fc0926865c6","login":"incakoala","name":"Vy","avatar_url":"https://avatars-next.observableusercontent.com/5d008ae093fa8b024defa3485b161456bc6ba621b7c21ef65061c11f7ca7dd98","tier":"starter_2024"},"version":1131,"publish_version":1131,"role":null,"authored":false,"authors":[{"type":"individual","id":"ded1682d7fdbac61","login":"incakoala","name":"Vy","avatar_url":"https://avatars-next.observableusercontent.com/5d008ae093fa8b024defa3485b161456bc6ba621b7c21ef65061c11f7ca7dd98","tier":null}],"members":[{"type":"individual","id":"ded1682d7fdbac61","login":"incakoala","name":"Vy","avatar_url":"https://avatars-next.observableusercontent.com/5d008ae093fa8b024defa3485b161456bc6ba621b7c21ef65061c11f7ca7dd98","tier":null,"role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":2,"collection":null,"type":"notebook","body":{"title":"🌳 Tree inequity: The wealthier and whiter a neighborhood is, the greener it gets","cells":[{"id":0,"value":"# 🌳 Tree inequity: The wealthier and whiter a neighborhood is, the greener it gets\n### A closer look into Rhode Island's Tree Equity data \n\n* *Part of NEU's [Data Visualization course](https://johnguerra.co/classes/data_visualization_spring_2022/) by John Alexis Guerra Gómez*\n* [Data source](https://www.treeequityscore.org/data/)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":792,"value":"$34","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":545,"value":"$35","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":798,"value":"$36","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":873,"value":"## Tasks\n\n### Task 1. 💰 Median Household Income vs 🌳 Tree Canopy\n * **T1.1.** Identify the correlation between the median household income and tree canopy % of neighborhoods in Rhode Island **(Identify, Correlation)**\n * Marks:\n * Points: represents a census block group\n * Line: illustrates the linear regression of the correlation\n * Channels:\n * Vertical & Horizontal positions: represents the correlation between two attributes, as well as the strength of the correlation at different value regions\n * **T1.2.** Compare the average tree canopy % that made up the neighborhoods, broken down by neighborhoods within the same buckets of median household income **(Compare, Distribution)**\n * Marks:\n * Lines: represents median household income groups\n * Channels:\n * Horizontal position: represents the length of the lines (the longer the line, the more tree canopy one group of income has)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":779,"value":"$37","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":566,"value":"{\n const base = vl\n .markText({ align: \"right\" })\n .encode(\n vl.y().fieldN(\"medhhinc\").title(\"Median Household Income\").sort(\"-x\"),\n vl.x().average(\"treecanopy\").axis({ format: \"%\" }).title(\"Average Tree Canopy %\"),\n vl.text().average(\"treecanopy\").format(\".0%\")\n )\n .title({ text: 'Neighborhoods with more than $100k median income has at least 21% more tree coverage than lower-income neighborhoods', anchor: \"start\" })\n\n const bars = base\n .markBar({tooltip: true})\n .encode(vl.color({ value: \"#d4dc9a\" }))\n // .encode(vl.color({legend: null}).fieldN(\"medhhinc\").scale({scheme: 'redyellowgreen'}).sort(\"x\"))\n\n return vl\n .layer(bars, base)\n .data(data_incomes_grouped)\n .width(400)\n .height(200)\n .render();\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":998,"value":"$38","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":877,"value":"### Task 2. 👥 People of Color vs. 🌳 Tree Canopy\n * **T2.1.** Identify the correlation between the people of color % and tree canopy % of neighborhoods in Rhode Island **(Identify, Correlation)**\n * Marks:\n * Points: represents a census block group\n * Line: illustrates the linear regression of the correlation\n * Channels:\n * Vertical & Horizontal positions: represents the correlation between two attributes, as well as the strength of the correlation at different value regions\n * **T2.2.** Compare the average tree canopy % that made up the neighborhoods, broken down by grouped percentages of people of color **(Compare, Distribution)**\n * Marks:\n * Lines: represents people of color percentage groups\n * Channels:\n * Horizontal position: represents the length of the lines (the longer the line, the more tree canopy one group has)","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":279,"value":"$39","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":723,"value":"{\n const base = vl\n .markText({ align: \"right\" })\n .encode(\n vl.y().fieldN(\"pctpoc\").title(\"People of Color %\"),\n vl.x().average(\"treecanopy\").axis({ format: \"%\" }).title(\"Average Tree Canopy %\"),\n vl.text().average(\"treecanopy\").format(\".0%\")\n )\n .title({ text: 'Trees cover almost half of the surface in predominantly white neighborhoods, compared to barely one-fourth in communities of color', anchor: \"start\" })\n\n const bars = base\n .markBar({tooltip: true})\n .encode(vl.color({ value: \"#d4dc9a\" }))\n // .encode(vl.color({legend: null}).fieldN(\"pctpoc\").scale({scheme: 'redyellowgreen'}).sort(\"-x\"))\n\n return vl\n .layer(bars, base)\n .data(data_poc_grouped)\n .width(600)\n .height(300)\n .render();\n}","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":818,"value":"$3a","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":795,"value":"## Data Processing\n\nThe data was downloaded from the source, and the original .dbf file was converted to .csv using this free [online tool](https://onlineconvertfree.com/convert-format/dbf-to-csv/).\n\nIn order to represent the data nicely on the bar charts, I created 2 new arrays `data_incomes_grouped` and `data_poc_grouped` which aggregate the raw data into groups of medium household income and people of color %, respectively.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"","database":"$undefined"},{"id":584,"value":"data_incomes_grouped = raw_data.reduce((acc, curr) => {\n let obj = {};\n var income = curr.medhhinc;\n if (income < 30000) {\n obj.medhhinc = \"0-$30k\";\n }\n else if (30000 <= income && income < 45000){\n obj.medhhinc = \"$30k-$45k\";\n }\n else if (45000 <= income && income < 60000) {\n obj.medhhinc = \"$45k-$60k\";\n }\n else if (60000 <= income && income < 100000) {\n obj.medhhinc = \"$60k-$100k\";\n }\n else if (100000 <= income && income < 250000) {\n obj.medhhinc = \"$100k-$250k\";\n }\n\n obj.treecanopy = curr.treecanopy;\n acc.push(obj)\n return acc;\n}, [])","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":719,"value":"data_poc_grouped = raw_data.reduce((acc, curr) => {\n let obj = {};\n var poc = curr.pctpoc;\n if (poc < 0.1) {\n obj.pctpoc = \"0-10%\";\n }\n else if (0.1 <= poc && poc < 0.2){\n obj.pctpoc = \"10-20%\";\n }\n else if (0.2 <= poc && poc < 0.3){\n obj.pctpoc = \"20-30%\";\n }\n else if (0.3 <= poc && poc < 0.4){\n obj.pctpoc = \"30-40%\";\n }\n else if (0.4 <= poc && poc < 0.5){\n obj.pctpoc = \"40-50%\";\n }\n else if (0.5 <= poc && poc < 0.6){\n obj.pctpoc = \"50-60%\";\n }\n else if (0.6 <= poc && poc < 0.7){\n obj.pctpoc = \"60-70%\";\n }\n else if (0.7 <= poc && poc < 0.8){\n obj.pctpoc = \"70-80%\";\n }\n else if (0.8 <= poc && poc < 0.9){\n obj.pctpoc = \"80-90%\";\n }\n else if (0.9 <= poc && poc < 1){\n obj.pctpoc = \"90-100%\";\n }\n else {\n obj.pctpoc = \"Unknown\";\n }\n\n obj.treecanopy = curr.treecanopy;\n acc.push(obj)\n return acc;\n}, [])","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":24,"value":"raw_data = FileAttachment(\"ri@1.csv\").csv()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":512,"value":"viewof explore = navio(raw_data)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":516,"value":"Inputs.table(explore)","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":12,"value":"import { navio } from \"@john-guerra/navio\"","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":28,"value":"import { vl } from \"@vega/vega-lite-api-v5\"","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W3b","stdlib":"1","titling":"auto"}} 33:"$32:body"