Uncertainty Values - Revision history

Admin at 12:17, 10 March 2021

2021-03-10T12:17:32Z

Admin at 12:13, 10 March 2021

2021-03-10T12:13:28Z

Admin at 12:12, 10 March 2021

2021-03-10T12:12:57Z

Admin at 12:12, 10 March 2021

2021-03-10T12:12:22Z

Admin at 10:12, 12 November 2020

2020-11-12T10:12:54Z

Admin at 10:12, 12 November 2020

2020-11-12T10:12:34Z

Admin at 10:11, 12 November 2020

2020-11-12T10:11:58Z

Admin at 10:11, 12 November 2020

2020-11-12T10:11:12Z

Admin: Created page with " [Home] - [back to GFMS Products] ---- '''Aggregated uncertainty values''' are generated along with the binary map product as a simplified appraisal..."

2020-11-12T10:07:46Z

Created page with " [Home] - [back to GFMS Products] ---- '''Aggregated uncertainty values''' are generated along with the binary map product as a simplified appraisal..."

New page

[[GFMS | [Home]]] - [[PRODUCTS | [back to GFMS Products]]]
----

'''Aggregated uncertainty values''' are generated along with the binary map product as a simplified appraisal of trust in the ensemble flood extent detection approach. <br>
First, uncertainty information from each of the individual algorithms is expressed in the same numerical range [0, 100] to ensure comparability and to facilitate further harmonization. <br>
In particular, probabilistic values from the [[ALGORITHM#HASARD (LIST) | LIST]] algorithm and fuzzy membership values from the DLR algorithm are first converted into classification uncertainty and next multiplied by a factor of 100. <br>
For DLR algorithm, since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. The conversion of these probabilistic (resp. fuzzy membership) values is carried out as U = Wi * (100 - Pi) + (1 - Wi) * Pi where Wi represents the binary water map classification (0 for dry and 1 for water classes), U is the uncertainty value and Pi is the probabilistic (resp. fuzzy membership) value. <br>
TUW classification uncertainty values, ranging from 0 and 1000, are divided by 10 in order to fall with the range [0,100]. Eventually, the average of the uncertainties from the three algorithms is taken. This has a value in the range of [0, 100], where values towards 100 indicate high confidence in the ensemble flood extent detection approach. Consequently, uncertainty information provided along with the map product communicates how much confidence is associated with the S-1 classification. End users can then use highly certain flood map products to identify resource requirements over areas of flood exposure to make timely emergency response decisions.

[[File:figure6.jpg]]

----
[[GFMS | [Home]]] - [[PRODUCTS | [back to GFMS Products]]]

← Older revision		Revision as of 12:17, 10 March 2021
Line 1:		Line 1:
		+	[[File:example7.jpg\|right\|100px\|caption]]
	[[GFMS \| [Home]]] - [[PRODUCTS \| [back to GFM Products]]]		[[GFMS \| [Home]]] - [[PRODUCTS \| [back to GFM Products]]]
	----		----

@@ Line 15: / Line 15: @@
 * Pi is the probabilistic (resp. fuzzy membership) value. <br>
-[[ALGORITHMS#Sentinel-1 flood mapping (TU Wien) | TU Wien algorithm]] classification uncertainty values, ranging from 0 and 1000, are divided by 10 in order to fall with the range [0,100]. Eventually, the average of the uncertainties from the three algorithms is taken. <br>
+[[ALGORITHM3 | TU Wien algorithm]] classification uncertainty values, ranging from 0 and 1000, are divided by 10 in order to fall with the range [0,100]. Eventually, the average of the uncertainties from the three algorithms is taken. <br>
 This has a value in the range of [0, 100], where values towards 100 indicate high confidence in the ensemble flood extent detection approach. Consequently, uncertainty information provided along with the map product communicates how much confidence is associated with the S-1 classification. End users can then use highly certain flood map products to identify resource requirements over areas of flood exposure to make timely emergency response decisions.

@@ Line 7: / Line 7: @@
 In particular, probabilistic values from the [[ALGORITHMS#HASARD (LIST) | LIST algorithm]]  and fuzzy membership values from the DLR algorithm are first converted into classification uncertainty and next multiplied by a factor of 100. <br>
-For [[ALGORITHMS#S-1FS (DLR) | DLR algorithm]] , since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. <br>
+For [[ALGORITHM2 | DLR algorithm]] , since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. <br>
 The conversion of these probabilistic (resp. fuzzy membership) values is carried out as: <br>
 ''U = Wi * (100 - Pi) + (1 - Wi) * Pi'' <br>

@@ Line 1: / Line 1: @@
-[[GFMS | [Home]]] - [[PRODUCTS | [back to GFMS Products]]]
+[[GFMS | [Home]]] - [[PRODUCTS | [back to GFM Products]]]
 ----
@@ Line 19: / Line 19: @@
-[[File:figure6.jpg]]
+[[File:figure6.jpg|border|centre]]
 ----
-[[GFMS | [Home]]] - [[PRODUCTS | [back to GFMS Products]]]
+[[GFMS | [Home]]] - [[PRODUCTS | [back to GFM Products]]]

@@ Line 7: / Line 7: @@
 In particular, probabilistic values from the [[ALGORITHMS#HASARD (LIST) | LIST algorithm]]  and fuzzy membership values from the DLR algorithm are first converted into classification uncertainty and next multiplied by a factor of 100. <br>
-For [[ALGORITHMS#S-1FS (DLR) | DLR algorithm]] , since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. The conversion of these probabilistic (resp. fuzzy membership) values is carried out as <br>
+For [[ALGORITHMS#S-1FS (DLR) | DLR algorithm]] , since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. <br>
 ''U = Wi * (100 - Pi) + (1 - Wi) * Pi'' <br>
 where:

← Older revision		Revision as of 10:11, 12 November 2020
Line 4:		Line 4:
	'''Aggregated uncertainty values''' are generated along with the binary map product as a simplified appraisal of trust in the ensemble flood extent detection approach. <br>		'''Aggregated uncertainty values''' are generated along with the binary map product as a simplified appraisal of trust in the ensemble flood extent detection approach. <br>
	First, uncertainty information from each of the individual algorithms is expressed in the same numerical range [0, 100] to ensure comparability and to facilitate further harmonization. <br>		First, uncertainty information from each of the individual algorithms is expressed in the same numerical range [0, 100] to ensure comparability and to facilitate further harmonization. <br>
−	In particular, probabilistic values from the [[ALGORITHM#HASARD (LIST) \| LIST]] algorithm and fuzzy membership values from the DLR algorithm are first converted into classification uncertainty and next multiplied by a factor of 100. <br>
−	For DLR algorithm, since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. The conversion of these probabilistic (resp. fuzzy membership) values is carried out as U = Wi * (100 - Pi) + (1 - Wi) * Pi where Wi represents the binary water map classification (0 for dry and 1 for water classes), U is the uncertainty value and Pi is the probabilistic (resp. fuzzy membership) value. <br>
−	TUW classification uncertainty values, ranging from 0 and 1000, are divided by 10 in order to fall with the range [0,100]. Eventually, the average of the uncertainties from the three algorithms is taken. This has a value in the range of [0, 100], where values towards 100 indicate high confidence in the ensemble flood extent detection approach. Consequently, uncertainty information provided along with the map product communicates how much confidence is associated with the S-1 classification. End users can then use highly certain flood map products to identify resource requirements over areas of flood exposure to make timely emergency response decisions.
−
−
−
−

		+	In particular, probabilistic values from the [[ALGORITHMS#HASARD (LIST) \| LIST algorithm]] and fuzzy membership values from the DLR algorithm are first converted into classification uncertainty and next multiplied by a factor of 100. <br>

		+	For [[ALGORITHMS#S-1FS (DLR) \| DLR algorithm]] , since fuzzy membership values are only assigned to water pixels, all unflooded pixels are assigned a value of 0. The conversion of these probabilistic (resp. fuzzy membership) values is carried out as U = Wi * (100 - Pi) + (1 - Wi) * Pi where:
		+	* Wi represents the binary water map classification (0 for dry and 1 for water classes)
		+	* U is the uncertainty value
		+	* Pi is the probabilistic (resp. fuzzy membership) value. <br>

		+	[[ALGORITHMS#Sentinel-1 flood mapping (TU Wien) \| TU Wien algorithm]] classification uncertainty values, ranging from 0 and 1000, are divided by 10 in order to fall with the range [0,100]. Eventually, the average of the uncertainties from the three algorithms is taken. <br>
		+	This has a value in the range of [0, 100], where values towards 100 indicate high confidence in the ensemble flood extent detection approach. Consequently, uncertainty information provided along with the map product communicates how much confidence is associated with the S-1 classification. End users can then use highly certain flood map products to identify resource requirements over areas of flood exposure to make timely emergency response decisions.