-
Notifications
You must be signed in to change notification settings - Fork 0
/
2019-02_iab_france_ta.html
405 lines (285 loc) · 11.9 KB
/
2019-02_iab_france_ta.html
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
<!DOCTYPE html>
<html>
<head>
<title>A multi-resolution daily air temperature model for France from MODIS and Landsat thermal data</title>
<meta charset="utf-8">
<meta name="author" content="Ian Hough" />
<meta name="date" content="2019-02-15" />
<link href="libs/remark-css-0.0.1/default.css" rel="stylesheet" />
<link href="libs/remark-css-0.0.1/default-fonts.css" rel="stylesheet" />
<link rel="stylesheet" href="custom.css" type="text/css" />
</head>
<body>
<textarea id="source">
class: center, middle, inverse, title-slide
# A multi-resolution daily air temperature model for France from MODIS and Landsat thermal data
### Ian Hough
### 2019-02-15
---
# Climate change, air pollution, and perinatal health
### Ian Hough
.pull-left[
### Dr. Johanna Lepeule
Institute for Advanced Biosciences
University Grenoble Alpes
]
.pull-right[
### Dr. Itai Kloog
Department of Geography and Environmental Development
Ben Gurion University of the Negev
]
.logos[  ]
---
background-image: url(img/outline-black.png)
---
# Adverse birth outcomes
### Preterm birth (<37 weeks gestation)
* 11% of all births and increasing (Harrison, et al., 2016)
* Leading cause of child mortality (Liu, et al., 2016)
* Sequalae in childhood and adulthood (McCormick, et al., 2011)
- Asthma, cerebral palsy, behavioural problems, etc.
### Term low birth weight (<2500 g)
* Increased morbidity and mortality in childhood & adulthood (Barker, 2004; Belbasis, et al., 2016)
---
# Ambient temperature (T<sub>a</sub>) and health
* Heat, cold, or variable T<sub>a</sub> can increase risk (Zhang, et al., 2017)
* Response may depend on local population & climate
* Hard to synthesize findings
| | Preterm birth | Birth weight | Term low birth weight |
|-----------|-------------------|-------------------|-----------------------|
| Exposure | Cold (<10th %ile) | IQR Ta increase | Heat (>95th %ile) |
| Window | Weeks 1–7 | Last 30 days | Trimester 3 |
| Statistic | Relative risk | Decrease | Odds ratio |
| Effect | 1.09 [1.04–1.15] | 16.6 g [5.9–27.4] | 1.31 [1.15–1.49] |
| Reference | Ha, et al. (2017) | Kloog, et al. (2015) | Ha, et al. (2017) |
---
class: inverse center middle
# How do we estimate T<sub>a</sub> exposure?
---
background-image: url(img/map_stations.jpg)
background-size: contain
class: inverse
---
background-image: url(img/map_stations_zoom.jpg)
background-size: contain
class: inverse
---
# Exposure error
* ### Sparse monitoring networks
* ### Coarse gridded meteorological data
### &rarr; May bias effect estimates towards null
---
# Our T<sub>a</sub> model
* ### Daily minimum, maximum, and mean T<sub>a</sub> 2000 - 2016
* ### 1 x 1 km<sup>2</sup> for continental France<sup>1</sup>
* ### 200 x 200 m<sup>2</sup> for large urban areas
.footnote[
[1] Extension of (Kloog, et al., 2017) (daily 1 km mean T<sub>a</sub> 2000 - 2011)
]
---
# Model components
### 1. Spatiotemporal and spatial predictors
* Land Surface Temperature (LST), elevation, etc.
### 2. Linear mixed model
* T<sub>a</sub> ~ LST with daily varying slope
### 3. Gapfilling
* T<sub>pred</sub> ~ T<sub>a</sub> at nearby stations
### 4. Local interpolation of residuals
* High spatial resolution predictors + machine learning ensemble
---
# Satellite data
### MODIS (1 km)
* Land Surface Temperature (LST)
* Terra: 10:30 / 22:30 (day / night)
* Aqua: 13:30 / 01:30 (day / night)
* NDVI
* Monthly composite
### Landsat 5 / 7 / 8 (30 m)
* Top-of-atmosphere brightness temperature (T<sub>B</sub>)
* NDVI
* &uarr; composited by month across 2000 - 2016
---
# Spatial predictors
* ### Elevation
* ### Land cover
* ### Population
* ### Climatic regions
&uarr; Aggregated to 1 km and 200 m grids
---
# Stage 1: linear mixed model (1 km)
`$$T_a = (\alpha + \mu_{jr}) + (\beta_1 + \nu_{jr}) \cdot LST + \beta_2 \cdot Emissivity + \\
\; \; \; \; \; \; \; \; \beta_3 \cdot NDVI + \beta_4 \cdot Elevation + \beta_5 \cdot Population + \\
\; \; \; \; \; \; \; \; \beta_6 \cdot Land Cover + e$$`
*j* = day &nbsp; *r* = climatic region &nbsp; *e* = error
---
# Stage 2: Gapfilling
`$$T_{pred} = (\alpha + \mu_{ip}) + (\beta_1 + \nu_{ip}) \cdot T_{IDW} + e$$`
*i* = grid cell
*p* = two-month period
*T<sub>IDW</sub>* = inverse distance weighted T<sub>a</sub>
---
background-image: url(img/map_ta_1km.jpg)
background-size: contain
class: inverse
---
# 1 km model performance
### Cross-validated 1 km predictions (calibration stage)
| 2000-2016 | R2 | RMSE | MAE | Spatial R2 | Spatial RMSE | Temporal R2 | Temporal RMSE |
|--------------------|------|------|-----|------------|--------------|-------------|---------------|
| T<sub>a</sub> min | 0.92 | 1.9 | 1.4 | 0.89 | 1.1 | 0.94 | 1.6 |
| T<sub>a</sub> mean | 0.97 | 1.3 | 0.9 | 0.95 | 0.8 | 0.97 | 1.2 |
| T<sub>a</sub> max | 0.95 | 1.8 | 1.4 | 0.88 | 1.2 | 0.96 | 1.5 |
### Previous model (Kloog, et al., 2017)
| 2000-2011 | R2 | RMSE | MAE | Spatial R2 | Spatial RMSE | Temporal R2 | Temporal RMSE |
|--------------------|------|------|-----|------------|--------------|-------------|---------------|
| T<sub>a</sub> mean | 0.95 | 1.5 | * | 0.91 | 0.65 | 0.96 | * |
\* = not reported
---
background-image: url(img/map_stations_zoom.jpg)
background-size: contain
class: inverse
---
background-image: url(img/map_ta_1km_zoom.jpg)
background-size: contain
class: inverse
---
# Stage 3: Residual interpolation (200 m)
### Contiguous urban areas with > 50,000 inhabitants
### Random forest and XGBoost models
`$$R \sim T_{pred},\ T_{B},\ NDVI,\ Elevation,\ Population,\\Land Cover,\ lat,\ lon,\ day$$`
### GAM ensemble
* Weights vary by location and predicted residual
---
background-image: url(img/map_ta_1km_zoom.jpg)
background-size: contain
class: inverse
---
background-image: url(img/map_ta_200m.jpg)
background-size: contain
class: inverse
---
# 200 m model performance
### Cross-validated 200 m ensemble predictions (residual scale)
| 2000-2016 | R2 | RMSE | MAE | Spatial R2 | Spatial RMSE | Temporal R2 | Temporal RMSE |
|------------------|------|------|------|------------|--------------|-------------|---------------|
| R<sub>min</sub> | 0.79 | 0.6 | 0.4 | 1.0 | 0.05 | 0.66 | 0.6 |
| R<sub>mean</sub> | 0.89 | 0.4 | 0.3 | 1.0 | 0.04 | 0.87 | 0.4 |
| R<sub>max</sub> | 0.85 | 0.5 | 0.3 | 1.0 | 0.03 | 0.73 | 0.5 |
---
# Next steps
### Fine particulate matter models (PM<sub>10</sub> & PM<sub>2.5</sub>)
* Similar to T<sub>a</sub> model
* MODIS aerosol optical depth (AOD)
### Birth outcomes study
* EDEN, PELAGIE, SEPAGES
* Birth weight and preterm birth
* T<sub>a</sub>, PM, and interaction
---
class: inverse middle center
# Thanks!
.footnote[.left[
### Ian Hough
]]
---
class: inverse references-slide
# References
Barker, D. J. P. (2004). "The Developmental Origins of Adult
Disease". In: _Journal of the American College of Nutrition_
23.2004, pp. 588S-595S. DOI: 10.1080/07315724.2004.10719428.
Belbasis, L, et al. (2016). "Birth weight in relation to health
and disease in later life: An umbrella review of systematic
reviews and meta-analyses". In: _BMC Medicine_ 14. DOI:
10.1186/s12916-016-0692-5.
Ha, S, et al. (2017). "Ambient temperature and early delivery of
Singleton Pregnancies". In: _Environmental Health Perspectives_
125.3, pp. 453-459. DOI: 10.1289/EHP97.
Ha, S, et al. (2017). "Ambient temperature and air quality in
relation to small for gestational age and term low birthweight".
In: _Environmental Research_ 155, pp. 394-400. DOI:
10.1016/j.envres.2017.02.021.
Harrison, M. S, et al. (2016). "Global burden of prematurity". In:
_Seminars in Fetal and Neonatal Medicine_ 21.2, pp. 74-79. DOI:
10.1016/j.siny.2015.12.007.
Kloog, I, et al. (2014). "A new hybrid spatio-temporal model for
estimating daily multi-year PM2.5 concentrations across
northeastern USA using high resolution aerosol optical depth
data". 95, pp. 581-590. DOI: 10.1016/j.atmosenv.2014.07.014.
Kloog, I, et al. (2015). "Using Satellite-Based Spatiotemporal
Resolved Air Temperature Exposure to Study the Association between
Ambient Air Temperature and Birth Outcomes in Massachusetts". In:
_Environmental Health Perspectives_ 123.10, pp. 1053-1058. DOI:
10.1289/ehp.1308075.
Kloog, I, et al. (2017). "Modelling spatio-temporally resolved air
temperature across the complex geo-climate area of France using
satellite-derived land surface temperature data". In:
_International Journal of Climatology_ 37.1, pp. 296-304. DOI:
10.1002/joc.4705.
Liu, L, et al. (2016). "Global, regional, and national causes of
under-5 mortality in 2000-15: an updated systematic analysis with
implications for the Sustainable Development Goals". In: _The
Lancet_ 388, pp. 3027-3035. DOI: 10.1016/S0140-6736(16)31593-8.
McCormick, M. C, et al. (2011). "Prematurity: An Overview and
Public Health Implications". In: _Annual Review of Public Health_
32, pp. 367-379. DOI: 10.1146/annurev-publhealth-090810-182459.
Rosenfeld, A, et al. (2017). "Estimating daily minimum, maximum,
and mean near surface air temperature using hybrid satellite
models across Israel". In: _Environmental Research_ 159, pp.
297-312. DOI: 10.1016/j.envres.2017.08.017.
Shi, L, et al. (2016). "Estimating daily air temperature across
the Southeastern United States using high-resolution satellite
data: A statistical modeling study". In: _Environmental Research_
146, pp. 51-58. DOI: 10.1016/j.envres.2015.12.006.
Zhang, Y, et al. (2017). "Temperature exposure during pregnancy
and birth outcomes: An updated systematic review of
epidemiological evidence". In: _Environmental Pollution_ 225, pp.
700-712. DOI: 10.1016/j.envpol.2017.02.066.
</textarea>
<script src="https://remarkjs.com/downloads/remark-latest.min.js"></script>
<script>var slideshow = remark.create({
"highlightStyle": "github",
"highlightLines": true,
"countIncrementalSlides": false
});
if (window.HTMLWidgets) slideshow.on('afterShowSlide', function (slide) {
window.dispatchEvent(new Event('resize'));
});
(function() {
var d = document, s = d.createElement("style"), r = d.querySelector(".remark-slide-scaler");
if (!r) return;
s.type = "text/css"; s.innerHTML = "@page {size: " + r.style.width + " " + r.style.height +"; }";
d.head.appendChild(s);
})();</script>
<script>
(function() {
var i, text, code, codes = document.getElementsByTagName('code');
for (i = 0; i < codes.length;) {
code = codes[i];
if (code.parentNode.tagName !== 'PRE' && code.childElementCount === 0) {
text = code.textContent;
if (/^\\\((.|\s)+\\\)$/.test(text) || /^\\\[(.|\s)+\\\]$/.test(text) ||
/^\$\$(.|\s)+\$\$$/.test(text) ||
/^\\begin\{([^}]+)\}(.|\s)+\\end\{[^}]+\}$/.test(text)) {
code.outerHTML = code.innerHTML; // remove <code></code>
continue;
}
}
i++;
}
})();
</script>
<!-- dynamically load mathjax for compatibility with self-contained -->
<script>
(function () {
var script = document.createElement('script');
script.type = 'text/javascript';
script.src = 'https://mathjax.rstudio.com/latest/MathJax.js?config=TeX-MML-AM_CHTML';
if (location.protocol !== 'file:' && /^https?:/.test(script.src))
script.src = script.src.replace(/^https?:/, '');
document.getElementsByTagName('head')[0].appendChild(script);
})();
</script>
</body>
</html>