-
Notifications
You must be signed in to change notification settings - Fork 0
/
wind_enviro_monitoring
294 lines (259 loc) · 22 KB
/
wind_enviro_monitoring
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
---
maths : true
gestionGrosMots: true
rechercheContenu: true
clavier: false
---
# Chatbot : About environmental monitoring of onshore wind farms in France
> **<span class="unique">I am a Chatbot [ChatMD](https://eyssette.forge.aeif.fr/chatMD) developped by [Guillaume Marchais](https://www.linkedin.com/in/gmarchais) who compiles various publically-made sources that he did not produced himself</span>**
>
> I can help you understand the scientific aspects of this issue.
>
> What do you want know? _**(scroll down for more)**_
>
> 1. [Which species are more sensitive to collisions with wind turbines](esp sensibles)
> 2. [What is the national protocol for post-construction monitoring in France ?](Le protocole national 2018)
> 3. [What is the "roads & pads" protocol](Le protocole allégé roads & pads)
> 4. [About the searcher efficiency test](Le test de détection)
> 5. [About the persistence test](Le test de persistance)
> 6. [About the area correction](La correction surfacique)
> 7. [The statistical models for the estimation of mortality](Les formules d'estimation)
> 8. [The acoustic monitoring of bat activity at nacelle height](suivi chiros)
> 9. [The bird activity monitoring](suivi ornitho)
## esp sensibles
- sensitive species
- bird collisions
- bat collisions
Birds of all species can collide with wind turbines, but those that reproduce in small numbers are particularly vulnerable. This includes most raptors, which have large territories and are often found in small numbers, even during migration. Other species that are at risk include gulls and terns, which are found near coasts, and cranes and storks, which are less common.
Migratory bats are also particurlarly vulnerable such as the ([Nathusius' Pipistrelle ](https://youtu.be/ImYYqXYfTl0?si=AdOOy4fumnpiI9bx) in Europe.
More ?
1. [Yes](Durr)
2. [No](Non fin)
## Durr
- Durr
- Dürr
A German scientist, Tobias Dürr, has been collecting data on bird mortality at wind turbines in Europe for many years. This data is not complete, but it provides valuable insights into the species that are most at risk from wind turbines. The website [Landesamt für Umwelt Brandenburg](https://lfu.brandenburg.de/lfu/de/aufgaben/natur/artenschutz/vogelschutzwarte/arbeitsschwerpunkt-entwicklung-und-umsetzung-von-schutzstrategien/auswirkungen-von-windenergieanlagen-auf-voegel-und-fledermaeuse/) makes data avaliable for download as Excel spreadsheets (at the bottom of the page).
Beyond the simple number of birds killed, it is important to consider the mortality rates relative to the populations of each species and their conservation status. This is because the impact of wind turbines on bird populations will vary depending on the size of the population, the number of wind turbines, and the conservation status of the species. This is issue is discussed in the article of the French journal "Eolien & Biodiversité" de 2022 (page 23) => [https://www.calameo.com/books/0063425990f6eb7c24b96](https://www.calameo.com/books/0063425990f6eb7c24b96)
More ?
1. [Yes]()
2. [No](Non fin)
## Le protocole national 2018
- French protocole
- French national protocole 2018
- standardised protocole in France
- protocole France
In France, a minimum set of practices for the research of birds and bats impacted by wind turbines has been established and validated. The first version of the protocol was published in 2015, and a more complete version was published in 2018. The second version is available for download [here](https://eolien-biodiversite.com/IMG/pdf/protocole_de_suivi_revision_2018.pdf).
The main requirements are as follows: :
* **The number of visits**
* at least 20 visits per year
* between mid-May and the end of October (from week 20 to 43)
* **The areas to search for carcasses**
* A 100 m x 100 m square or concentric circles with a radius equal to the length of the blade (minimum 50 m)
* Transect lines spaced 5 to 10 m apart, depending on the condition of the ground and vegetation
* **The tests**
* a test enabling to assess the seacher efficiency
* a test enabling to assess the local persistence of the carcasses (scavenger presence and farming practices)
* Continuous recording of bat activity at nacelle height for a large part of the season
**The objective is to determine the real mortality rates by using statistical models to estimate them from the observed raw mortality on the ground.**
More?
1. [Yes]()
2. [No](Non fin)
## Les formules d'estimation
- estimation
- models
- statistical
- formula
Which one do you want to see :
1. [Huso](formule : Huso)
2. [Erickson](formule : Erickson)
3. [Dalthorp](formule : Dalthorp)
4. [Jones](formule : Jones)
## formule : Huso
- Huso
Huso et al. (2010) consider a mortality globally constant but with their approach the probability of removal at the middle of the interval between two consecutive visits is not equal to the average probability of persistence of a carcass.
$$N = \frac{C}{d * p * ê} * A$$ with $$p = t * (1-exp^-I/l) / l$$
even if the duration of persistence follows a negative exponantiel function.
Nevertheless, the Huso models are based on north american data where persistence is usually longer than found in Europe.
The biaises are therefore more important with this approach when peristence is short (a few days only).
Reference : [Huso et al. 2011](https://onlinelibrary.wiley.com/doi/10.1002/env.1052)
Is that clear?
1. [Yes](suite formules)
2. [No](Non fin)
## formule : Jones
- Jones
The model from Jones et al. (2009) is based on the following hypothesis :
* the mortality rate is considered globally constant on the intervall of consecutive searches ;
* the duration of persistence of a carcass follows a negative exponantiel function, the persistence rate is calculated as Tm (average duration in days) ;
* the probability of the average removal on the complete interval (when animals can impacted at the beginning or the end of the interval) can be approximated by the probability of removal at the middle of the interval.
Jones uses the notion of the effective interval for calculating the corrective factor ê which corresponds to the ratio between the duration before that 99 % of carcasses are removed and the duration of the interval. Indeed, it is quite logic that when the interval is long then the persistance ratio gets closer to 0. The effective interval corresponds to the duration for which the persistance ratio is equal to 1 % (100 - 99 %).
The average effective interval, Î is equal to : - log(0,01) x tm. With I having the minimal value between I et Î, noted ê.
More?
1. [Yes](suite formules)
2. [No](Non fin)
## formule : Dalthorp
- Dalthorp
He is the main developer of this open source application [GenEst](https://www.usgs.gov/software/genest-a-generalized-estimator-mortality](https://www.usgs.gov/software/genest-a-generalized-estimator-mortality) with M. Huso
which uses the advanced statistical models to enable a tailored approach for calculating the estimatiosn of mortality depending on the various conditions of the fatality search protocol.
Available to download for the USGS website and Github, it runs on as R software package.
Since 2018, this is most advanced international tool when a minimal number of carcasses are found (usually > 5).
The user guide in English is downloadable from [here](https://www.usgs.gov/software/genest-a-generalized-estimator-mortality).
A webinar in English can be viewed on [Youtube](https://youtu.be/pSZDwiApigA?si=sqJF_Dv4HETafarp).
More?
1. [Yes](suite formules)
2. [No](Non fin)
## formule : Erickson
- Erickson
The Erickson's estimator (2000) used to be frequently used in France (previously to 2018). This test tends to underestimate the data for the short persistane rates but not for the long ones. This result is also considered due to the constant sampling effort. The formula takes into account the frequency of the visits but not the average time of intervals (time in-between visits). It considers an explonantial decrease of the removal time of carcasses which can be found during the visits following the first one without findings.
More?
1. [Yes](suite formules)
2. [No](Non fin)
## suite formules
- best
- simple
- easy
- all formulas
- no carcasses...
Basically, you would like to kow...
1. [Which one is the simplest or easiest?](formule : simple)
2. [Which one is the best?](formule : best)
3. [Shall I use them all?](formule : use)
4. [Which one to choose when there were few carcasses or even none?](formule : EoA)
## formule : simple
Erickson' formula is the simplest and the oldest one but often understimates as it does not use a complex model for the removal of carcasses.
For further details, I recommand the reading of this technical guidance by the [French lecturer Aurélien Besnard du CNRS-CEFE](https://shiny.cefe.cnrs.fr/eolapp/estimations_de_mortalites.pdf).
All clear?
1. [Yes](Got it)
2. [No](Non fin)
## formule : best
It has been shown that there is not only one relevant or valid formula for all cases (long or short persistance etc...). Huso's formula (2010) is often used by dfault but it alo has its limits of validity when few carcasses have been found. It is is recommended to start using the open source R Package "GenEst" which Huso has developped with Dalthorp et al. => [https://www.usgs.gov/software/genest-a-generalized-estimator-mortality](https://www.usgs.gov/software/genest-a-generalized-estimator-mortality)
[Voigt et al.2022](https://www.sciencedirect.com/science/article/pii/S2351989422001512): <blockquote class="is-size-5 is-size-6-mobile has-text-justified">"..., we use the ‘GenEst’ value as this methodological approach is currently the most widely used statistical approach in estimating bat casualties at wind turbines."</blockquote>.
For further details, I recommand the reading of this technical guidance by the [French lecturer Aurélien Besnard du CNRS-CEFE](https://shiny.cefe.cnrs.fr/eolapp/estimations_de_mortalites.pdf).
All clear?
1. [Yes](Got it. Now, I would like to watch a video tutorial on GenEst such as this one from WREN by [Manuela Husos USGS](https://youtu.be/pSZDwiApigA?si=nCoDOg6vFgrKci2w))
2. [No](Non fin)
## formule : use
The French national protocol (2018) is requesting to use Huso's formula with two others to be able to show the variability of the resultats (median and confidence intervals) owing the various conditions.
Le protocole national 2018 demande d"utiliser celle de Huso 2010 avec deux autres d'entre elles pour montrer la variabilité des résultats selon les conditions.
For further details, I recommand the reading of this technical guidance by the [French lecturer Aurélien Besnard du CNRS-CEFE](https://shiny.cefe.cnrs.fr/eolapp/estimations_de_mortalites.pdf).
In 2017, a simple and free online app was designed by the French lecturers from CNRS-CEFE, [EOLAPP](https://shiny.cefe.cnrs.fr/eolapp/). [Here is a an introductory video in French](https://youtu.be/k8VA7JqfjXA).
All clear?
1. [Yes](Got it)
2. [No](Le protocole national 2018)
## formule : EoA
In this particular case, the different fomrulas cannot make a conclusive estimation of the real mortality. Another approach is required : the one which questions the **_credibility_** of the different paramaters of the monitoring (rather than the probability or the level of confidence based on a posterior distribution). Huso and Dalthorp have developped another open source software package called _**Evidence of Absence**_ which can be downloaded from the [USGS website
](https://pubs.usgs.gov/publication/ds1055). It works on R 4.0 or more using various packages and this free stats programme ["JAGS"](https://sourceforge.net/projects/mcmc-jags/).
In addition, experts from the american ecological consultancy firm "Western Ecosystems Technology" have developed an extension which uses more covariables in order to adjust even more the estimates. More details available [here](https://www.biorxiv.org/content/10.1101/2020.01.21.914754v3.full.pdf)
Do you want to go further ?
1. [Watch a webinar on EoA with Manuelo Huso from USGS](https://youtu.be/xn8jD2TGQpY?si=Ql3N_WABpE4BM9JK&t=569)
2. [No](Non fin)
## Le test de persistance
- persistence
- persistence rate
- persistence test
- interval
This consists in leaving on site various ploys or trial carcasses (frozen mice or rats) on searched areas in order to assess the persistance in time at various periods of the year (May and September for instance). The French protocole 2018 requires to use at least 3 per wind turbine for each sesssion, but it is recommended to increase this number to cover all types of visibility. In France, the persistance is relatively short (< 3 days) in open field when the ploys are left at the same time ; which is in itself a bias compared to the real situtation (relatively randomly distributed in time for a specific period). It is thus relevant to use a result for each specific tested period when making the estimations.
[To go further](https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0067997)
Do you want see an example ?
1. [Yes](Je veux voir un exemple)
2. [No](Non fin)
## Je veux voir un exemple
See Fig 3 from this [published paper](https://www.sciencedirect.com/science/article/pii/S2351989422001512)
Doy want to know more?
1. [Yes]()
2. [No](Non fin)
## Le test de détection
A searcher's detection capacity varies a lot depending the visibility conditions, his/her experience, tireness and so on... This test uses trial carcasses or similar objects placed on site in order to check how much of them are found indeed during a standard visit. The results must be tailored to each searcher, visibility class, size class and period of the year. Bad scores, i.e. below 50 %, have to well analysed as they mean that the search is not effective enough or the visibility is too poor.
Do you want to learn more?
1. [Yes]()
2. [No](Non fin)
## La correction surfacique
The [French protocol 2018](https://eolien-biodiversite.com/IMG/pdf/protocole_de_suivi_revision_2018.pdf) requires to cover circle area with the radius corresponding to the leght of the blade. When there is too much crops (wheat, maize, beetroot...), it becomes complicated to search without damaging them. It is then generally admitted to avoid these areas. Presence of cattle can also lead to the same decision. The simplest value corresponds to the proportional searched areas; and more adjusted one is to work out this every 5 or 10 m radius concentric circles (Arnett 2006).
Do you want to know more ?
1. [Yes](corr surfa DWP)
2. [No](Non fin)
## corr surfa DWP
More recently, an in-depth calculation of this parameter involves the probabilistic density in function of the distance to the mast => see this [module "DWP" developped by the experts form USGS](https://www.usgs.gov/software/density-weighted-proportion-dwp-tool-suite)
Do you want to know more ?
1. [Yes]()
2. [No](Non fin)
## Le protocole allégé roads & pads
- roads
- pads
- protocol roads & pads
- simple protocol
The searches for carcasses are focused on areas with hig visibility only, which often are the platforms or pads and the access roads up to 120 m whenever possible. It usually takes less than 10 min to cover around 20 % of the standardised full plot (100 m square). These quick visits can also be carried out during other types of monitoring such as bird activity surveys or in alternance with standardises visits for all wind turbines in order to increase the frequency. This is particularly relevant for sites with very low persistence.
Is this clear ?
1. [Yes](J'ai compris)
2. [No](reference rp)
## reference rp
This published paper investigates the comparaison of the roads & pads protocol to other with simulations using various parameters : ["Comparing methods to estimate the proportion of turbine-induced bird and bat mortality in the search area under a road and pad search protocol."](https://link.springer.com/article/10.1007/s10651-020-00466-0)
Do you want to learn more?
1. [Yes]()
2. [No](Non fin)
## suivi chiros
- acoustic monitoring
- nacelle
- bats
- bat monitoring
In parallel of fatality searches, bat activity is also monitored at nacelle height in order to understand better in which weather conditions bats are active throughout the seasons. This has been useful to work out bat curtailment plans specifically for each site.
Do you want to see few examples ?
1. [Yes](suivi chiros 2)
2. [No](Non fin)
## suivi chiros 2
- Probat
- allemand
- infrarouge
- Etats-Unis
In Europe, the germans used this [applied science tool](https://www.probat.org/probat-7) produced the research programm RENEBAT.
**Verteilung der Fledermausaktivität im Rotorbereich einer Windenergieanlage, Grafik: Ulrike Eberius**
[Here is a video in English](https://youtu.be/pSZDwiApigA?si=nCoDOg6vFgrKci2w) from Leibniz-Institut für Zoo- und Wildtierforschung (IZW) explaining the impacts on batsof wind turbines located in woodlands
In France and more recently, academics from the Museum d'Histoire Naturelle de Paris have published this article [(Barré et al., 2023)](https://www.sciencedirect.com/science/article/abs/pii/S0048969723000190) going deeper into which covariables may influence bat activity at in order to elaborate tailored bat curailment algorithms.
.
In addition, this other research project named "OPRECH" has compiled all types of bat curtailments implemented in France from 2010 to 2020. Full report in French available [here](https://librairie.ademe.fr/energies-renouvelables-reseaux-et-stockage/6242-programme-oprech.html).
Do you want to learn more ?
1. [Yes]()
2. [No](Non fin)
## J'ai compris
Well done, now you know much more. You may want to watch this video tutorial in English on data analysis by [P. Rabie from WEST Inc](https://eyssette.forge.aeif.fr/chatMD/#https://github.com/guillaumarchais/ChatMD/blob/main/)
Do you want to learn how to interpret all these data ?
1. [Yes](interprétation mortalité)
2. [No](Non fin)
## interprétation mortalité
Several initiatives are currently undergoing to pull together large amount of results enabling comparisons betewwen various wind farms of each region or country. Here is a few examples:
* Germany : [Voigt et al. 2022](https://www.sciencedirect.com/science/article/pii/S2351989422001512)
* France > région Grand-Est : [CPIE Pays de Soulaines 2018](http://www.plume-de-naturalistes.fr/wp-content/uploads/2019/04/09_TERNOIS_03-2019_Impact-developpement-eolien_Plume3_195-212.pdf)
* France > région Grand-Ouest : [Ouest Am](https://www.genie-ecologique.fr/synthese-des-donnees-de-suivis-post-implantation-deoliennes-dans-le-grand-ouest)
* Spain > région sud : [Salguero et al. 2023](https://link.springer.com/article/10.1007/s10531-023-02674-z)
* USA : [Lloyd et al. 2023](https://link.springer.com/article/10.1007/s10531-023-02674-z)
Beyond the comparisons, **the main objective is to significantly reduce mortality through various solutions and measures that consist of optimizing wind turbine curtailments at risk periods and during critical situations (animals flying too close to the blades).** Quite basicly, this takes the form of the presence of "monitors" or sentinels in the field who can quickly trigger a shutdown of the turbines (with a mobile phone application or an emergency call for example).
Technological advances also allow the development of engineering solutions based on remote video and/or audio detection. The challenge is to detect the bird early enough to slow down the rotation of the blades using sophisticated algorithms; and bats with sensitive, weather-resistant microphones.
Whether it's with high-tech cameras, radars or microphones, engineers are trying to improve their automatic detection and analysis capabilities of their various systems. These are tested in real situations and verified through monitoring controlled by the State's decentralized services.
Do you want to learn more ?
1. [Yes]()
2. [No](Non fin)
## suivi ornitho
Bird monitoring surveys are carried out using binoculars, telescops and even laser range finders from various high-up spots in the vicinity of the wind energy facility.
.
Here are references of protocols currently used in France:
* [https://www.lpo.fr/la-lpo-en-actions/connaissance-des-especes-sauvages/suivis-ornithologiques](https://www.lpo.fr/la-lpo-en-actions/connaissance-des-especes-sauvages/suivis-ornithologiques=)
* [https://www.instinct-animal.fr/guide/ornithologie/inventaires-et-comptages-ornithologiques/](https://www.instinct-animal.fr/guide/ornithologie/inventaires-et-comptages-ornithologiques/=)
Do you wan to learn more?
1. [Yes]()
2. [No](Non fin)
## Non fin
Despite my attempts, it was not good enough. The next step involves reading the various quoted references... or you do not want to go further and shut this window ; )
In any case, I invite you to look into this fairly complexe issue but quite major in the context of climate chage and biodiversity crisis. You may find interesting to browse these other web sites:
* [Bats and wind](https://www.batsandwind.org)
* [Post-construction Bird and Bat Fatality Monitoring for Onshore Wind Energy Facilities in Emerging Market Countries](https://www.ifc.org/en/insights-reports/2023/bird-bat-fatality-monitoring-onshore-wind-energy-facilities)
* [Windexchange](https://windexchange.energy.gov/projects/birds)