I am pleased (more like absolutely flabbergasted) to announce that I got the
@doescience
Early Career Award for my proposal on improving
#urban
representation in Earth system models.
The staggering amount of support from folks at
@PNNLab
definitely loaded the dice in my favor.
This paper (), which I co-authored, has recently been shared by multiple
#climatechange
denier accounts on
#Twitter
as evidence of the impact of
#urban
#heat
islands on long-term observed
#warming
.
Except the paper does not show that. Let's discuss: [1/n]
Being introduced to
#EarthEngine
back in 2016 was a game changer.
In my first project, I created a
#global
surface
#urban
#heat
island database.
We just updated it to version 5 with estimates through 2020. Here is the web app to explore the data:
#EOChat
Thermal
#remotesensing
is poorly constrained for
#urban
use cases. Quantitative estimates are sensitive to the land surface
#temperature
(LST) retrieval algorithm used.
See spatial variability of LST for summer 2019 from two algorithms over Philadelphia.
#EarthEngine
#EOChat
π¨Just published in
#RemoteSensing
of
#Environment
(), we reconstruct spatiotemporally continuous
#Landsat
land surface temperature (LST) for 1985β2020 over 511
#cities
, finding a mean surface urban heat island (Is) growth rate of 0.156Β Β°C/decade. [1/n]
Our
#global
surface
#urban
#heat
island (SUHI) dataset is finally available through
@NASA
's SEDAC
@ciesin
:
Includes annual, summertime, and wintertime SUHI intensities for daytime and nighttime for over 10,000 urban extents throughout the world.
I recently updated the Beginner's Cookbook to
#EarthEngine
, intended to be an introductory tutorial to geospatial analysis:
Changes include an additional section on Reducers and easier to follow steps for the example scripts.
Note the large differences between what
#satellite
-derived land surface
#temperature
gives us and the true variability in mean radiant temperature within and across neighborhoods in Philadelphia.
Read the paper for more details:
#UrbanClimate
Got my copy of
#EarthEngine
@EEFAbook
on cloud-based
#remotesensing
; result of a massive community effort.
Also, papers are nice & all, but since the majority of my education has relied on books, feels awesome to see my name as a chapter author (on
#urban
#heatisland
) in print.
2023 was stupidly productive.
Got the DOE Early Career award+first NASA funding, published 15 papers (led 2 of them; wish this was higher), presented at 5 conferences (+universities & national labs), reviewed ~100 papers & 2 proposals, & added >800 citations.
I may have peaked.
Starting tomorrow,
@UN
programmes, funds, agencies and offices are gathering in
#Brindisi
to strengthen inter-agency cooperation on
#space
. New this year is a full day devoted to training on applications and demonstrations of
#technologies
, thanks to cooperation with
@CEOSdotORG
.
@brianmlucey
@NathanCAGray
The hungry hunter analogy is particularly cringe-inducing since you are in finance and your research involves sitting (and probably eating) in front of a computer.
There has been a plethora of studies that use
#satellite
-derived land surface
#temperature
(LST) to estimate the cooling efficiency of
#trees
, particularly
#urban
trees.
π’In our recent
#openaccess
paper in
#ERL
, we show that this approach leads to massive overestimations. [1/n]
I am pleased to announce that I will be joining the Atmospheric Sciences and Global Change Division of the Pacific Northwest National Lab as a Research Scientist this fall.
The discourse around
#heatstress
sucks both on
#Twitter
& in the
#media
.
It is hot and it will get hotter (human-caused
#ClimateChange
+El NiΓ±o has a role this year).
But all metrics that include
#temperature
are not same. See their magnitude & variability in this figure.
π§΅...
Have been exploring
#urban
roughness elements (building and vegetation height) at 30 m using
#Copernicus
COPDEM30 and FABDEM datasets on
#EarthEngine
. Really interesting and has a bunch of applications.
Code:
Thanks for ingesting these,
@samapriyaroy
In case anyone is using both products, the change in algorithm between
@ESA
#WorldCover2020
and
#WorldCover2021
will probably contribute more to 'urbanization' than any actual change between the years.
I used this
#EarthEngine
app () to compare.
#EOChat
There seem to be a confidence estimates for every building polygon (green polygons have a confidence score greater than equal to 0.9).
That is awesome.
I haven't played around enough with the data to comment on overall accuracy, but they provide precision scores per tile, which is very useful.
Good to see uncertainty estimates for these large-scale estimates from data producers.
#Google
#OpenBuildings
Last year,
@WHO
revised
#airquality
standards after 15 years.
So now, 99.58% of us live in areas with annual average PM2.5 values above the recommended 5 Β΅g/m3. π€·ββοΈ
I made an
#EarthEngine
#app
to explore how many times higher than that standard the actual concentrations are.
There are several
#urban
extent datasets based on various
#global
estimates of urbanization.
How does the choice of urban extent impact estimates of surface
#urban
#heat
island (SUHI)?
We answer this question in our recently published paper: [1/n]
Ignoring urban surface meteorology is the norm from a climatological perspective, but urban micro-climate is important for people living and working in cities. Need more urban surface observations as remotely sensed, derived, and modeled products perform quite poorly over cities.
Dr. Robin Tanamachi @tornatrix.bsky.social πͺπ‘
This is why we don't put surface observing stations in major cities: The buildings modify the air flow! Clearest demonstration I've yet seen of this concept. Source:
The Beginner's Cookbook for using
#EarthEngine
is now a community tutorial:
I originally designed this workshop module last year (thanks to
@ecoangelhsu
), and have since updated it for
#Yale
#GISDay
2018 and (upcoming) 2019 ().
Surface temperature inferred from satellites is view dependent, which can bias remotely sensed estimates of
#urban
#climate
.
π¨In our recent paper () in
#ScienceBulletin
, we map this 'urban thermal anisotropy' for >5500 cities & quantify these biases. 1/n
After several updates and improvements, our global database of urban heat islands (link to original paper: ) is now in the
#EarthEngine
data catalog ().
The data can also be explored here:
The "types of scientific paper" meme from
@xkcd
is now mostly dead.
So I made one for
#urban
#climate
research.
Here's to eventually publishing papers covering all these categories.
If you use artificial intelligence (
#AI
) and Earth observation (
#EO
) to study
#urban
#climate
, consider submitting your manuscript to our special issue in the International Journal of Applied Earth Observation and Geoinformation.
More info:
My
#PhD
supervisor sent me a copy of the latest edition of his book on Boundary-Layer Meteorology.
This edition has a new chapter on the polluted boundary layer, a small part of which has some findings from my dissertation (which was not on urban climateπ€·ββοΈ).
@YaleEnvironment
Another future
#urban
projection dataset just dropped (), this one considering fraction of urbanization at 1 km for more scenarios (
#SSP
-
#RCP
combos).
Will process it & add it to the heap that I was supposed to write a paper on from almost a year ago.
"Google Earth Engine open-source code for Land Surface Temperature estimation from the Landsat series" by
@ermida_sofia
#EarthEngine
#Landsat
Streaming now:
After spending way too much time visually inspecting the
@Esri
and
@ESA_EO
/
@VITO_RS_
~10m
#landcover
products, I compared the two more systematically for
#urban
(built-up) class on
#EarthEngine
. Results for a small selection of countries; one product with >double urban fraction:
In our recently published paper in Remote Sensing of Environment (), we test the sensitivities of
#global
and regional (by climate zone)
#Landsat
-derived surface
#urban
heat island (SUHI) intensities to different (5) methods of prescribing land emissivity.
Just came across a recent paper on detection of surface
#thermal
discharge from global
#nuclear
power plants using
#Landsat
data:
That's a really cool use of
#satellite
-derived surface
#temperature
.
Image below shows occurrence of these thermal plumes.
UPS sent me an email 2 days ago about an incoming shipment. I was pulling my hair out trying to figure out what it was since I didn't order anything.
Turns out it was my plaque for being a
#Google
Developer Expert for
#EarthEngine
.
A hugeπto the
#GDE
program!
#officialGDE
Last year, we designed a web application () using
#GoogleEarthEngine
to characterize global urban heat islands, increasing traction of the accompanying paper ()
More information and a list of updates here:
Papers on
#urban
#heat
keep stating that
#Landsat
's
#thermal
data are at 30 m res.
Native res of thermal band, depending on generation, is between 60 m & 120 m. Then it is resampled to 30 m.
Given discontinuities at that scale over cities, there is little real signal at 30 m.
As others have noted, new
@ESRI
product does not seem to be capturing much at that 10 m res, especially evident around urban areas. Example over New Haven, CT vs
#NLCD
2016. I thought it would pick up New Haven Green at 10 m res; which shows up as developed open space in NLCD.
Breaking:Β
@Esri
Β launches 2020 global 10m land cover map powered byΒ Impact ObservatoryΒ deep learning algorithms andΒ CopernicusΒ European Space Agency - ESA Space AgencyΒ Sentinel-2 imagery, leveraging theΒ MicrosoftΒ
#AI
for Earth
#PlanetaryComputer
.
#eochat
@mariopansera
Pretty bad take.
1. x papers published this year could be the result of several years of work since peer review timelines vary.
2. Whether 'it' is humanly possible or not depends on what 'it' is. Fields and even research groups have different criteria for authorship.
It would be absolutely great if the
#GIS
portals of
#city
governments made the shapefiles of their cities available for download instead of sharing maps in only image and pdf formats.
My colleagues and I are guest editing a focus issue for
@IOPenvironment
on 'Human-Earth System Interactions under Climate Change'.
If you are working on something relevant to this broad topic, we encourage you to submit your work.
More details here:
Our old paper on a
#global
surface
#urban
#heatisland
database is apparently the most cited article published in this journal since 2019 (arbitrary threshold?).
I have several thoughts about this work now (more in the future), but am glad that it has garnered interest.
#EOchat
With
@EPA
reducing annual PM2.5 limit to 9 ug/m3, I wanted to share this app () that visualizes PM2.5 globally.
Note that this uses the recently revised
@WHO
#airquality
standard of 5 ug/m3.
99.58% of us live in areas with PM2.5 above the WHO threshold.
Playing around with
@googleearth
Dynamic World dataset to understand its use for delineating
#urban
areas. Random highly urban pixel over Tokyo (mosaiced probability close to 0.75 since 2016).
Interesting anomalies in the time series though.
#EarthEngine
#geospatial
#eochat
Three decades (1985 to 2015) of human settlement evolution over
#Shanghai
and other major
#urban
clusters in the Yangtze River Delta at ~30 m resolution.
Lighter shades show earlier settlements.
#EarthEngine
#ESA
#EOChat
Been in Tri Cities, Washington for over a year and haven't examined any satellite data over the region.
So here is normalized difference vegetation index (NDVI) for summer 2020 from
#Landsat
8.
Those circular fields are apparently the result of center-pivot
#irrigation
.
#pnw
Here is an
#EarthEngine
#webapp
I developed to explore
#global
distribution of particulate matter
#airpollution
:
Annual values are around 20 times the revised
@WHO
standard in many parts of the Indo-Gangetic Basin, where over half a billion people live.
Indiaβs entire population lives in areas where the air pollution exceeds
@WHO
standards. Almost seven in 10 Indians live where air pollution exceeds the countryβs own average acceptable level.
#airpollution
@UChiAir
#lifeexpectancy
Link to the manuscript (reach out to me if you do not have access):
I also developed an
#EarhEngine
web application to visualize the dataset (which I am currently updating):
I am bringing this back as things have gotten significantly worse (and I still haven't published in all these categories).
Also, QT with a meme you made.
You have seen me rant about
#satellite
-derived land surface
#temperature
on Twitter in the past.
If you are at
#ICUC11Sydney
, you can watch me do that IRL tomorrow morning in Mathews Room 231.
Seeing that title, I may have crammed too many subtopics into this talk.π€
Next up,
@TC_Chakraborty
takes the stage sharing how land surface temperature from satellites relates to ambient heat stress and air temperature within cities, based on crowdsourced ground-based data.
Our newly developed urban heat island database for the US suggests large disparities across income and racial groups:
The data can be visualized on this
#GoogleEarthEngine
app:
In our paper published in Advances in Atmospheric Sciences (), we review how
#urbanization
can impact regional climate and extreme weather, starting with urban
#heat
islands (canopy, surface, boundary-layer, etc.) and heat waves, to... [1/3]
@WMO
Most of these points are already covered in the original
#paper
(either in the main text or in the supplementary text).
Would be nice if people actually read the papers that they were using to support their own biases. [18/n]
Introducing the Auto Arborist Dataset, a multiview urban tree classification dataset that consists of ~2.6M trees and >320 genera, which can aid in the development of models for urban forest monitoring. Check it out at β
I will be talking about 'Examining Urban Heat Across Scales: From Observations to Model Simulations' this Monday (Sep 18th) for the
@michigantech
Environmental Engineering Graduate Seminar.
Event and streaming link:
#urban
#heat
#climate
#heatstress
I have a couple of talks on
#UrbanClimate
at
#ICUC11
next week. Also my first time attending this conference (I barely had a research career during ICUC10).
Since I have a ~24 hr flight to Sydney, I am hoping that is enough time to start & finish working on my slides.π€
Customized your
#EarthEngine
basemap. Learn how in this new
#community
contributed tutorial (congrats TC).πΊοΈπ
π
Have an idea for an EE tutorial? Here's how to submit it:
π
Brightness temperature (BT) is NOT land surface temperature (LST).
BT is the temperature of a hypothetical black body with the same emitted radiation as a surface. This has to be corrected using the emissivity of that surface to get LST, which is higher than BT for real cases.
π«Disinformation!π«
Land surface temperatures or "Brightness Temperatures" from satellites are NOT the same as air temperature that a thermometer would read.
On our recently published study on moist
#heatstress
disparities across
#US
cities.
Here we examine associations between residential
#segregation
(both present-day and past [
#redlining
]) on different measures of
#urban
#heat
exposure.
Link to paper:
Pretty much everywhere, itβs gonna be hotβand in U.S. cities, some people endure more heat than others.
New PNNL research finds that the average Black urban resident is exposed to disproportionally higher heat stress:
Great panel!
Mentioned this new paper on the intersection between human
#mobility
&
#urban
#heat
:
We often assume that people are exposed to the environment of their home census tract/block/etc., which is not always true.
Hope to see more data on this.
I will be talking about 'Isolating local urban climate signals and their relevance for public health' at the 'Climate Vulnerability Mapping - Available Tools and Research Gaps' session tomorrow, part of the CAFΓ Climate and Health Conference 2024.
#climatehealthcafe
Change in built-up pixels in and around Shanghai, China from 1992 to 2018.
Data Source:
#ESA
CCI land cover data
Platform:
#GoogleEarthEngine
PS: Took way too much work to get a base map to show with the timelapse.
#AGU23
abstract submission season has begun and I wanted to add another deadline (specifically, 2nd August) to your calendar.
I am co-convening 3 sessions this year, including 2 newly proposed sessions.
Looking forward to seeing the great
#science
everyone is doing.
Detailsβ¬οΈ
In our recent paper in
#AGU
#GRL
() , we show large differences in long-term trends in surface (from land surface
#temperature
or LST) & canopy (from air temperature or AT)
#urban
#heat
islands (UHIs) & their drivers across
#global
#cities
(n=5643). 1/n
Since I started convening AGU sessions, it has been fun to monitor how 80% of the submissions happen on the day of & the day before the deadline.
Reminds me of a quotation from Douglas Adams: βI love deadlines. I love the whooshing noise they make as they go by.β
#AGU23
MDPI definitely shows predatory tendencies (and so do many other for-profit publishers).
But Predatory Reports, the site posting this list, has 80 followers on Twitter and their about us page provides as much clarity as some of MDPI's editorial decisions.
Something to consider.