18 Bridging the Information Gap: Mapping Data Sets on Information …
219
Table 18.2 (continued)
Crisis impact
Operational environment
Flood duration
Security and access
Earlier predictions
News
Time of inundation
Accessibility
Inundated area
Security
Drainage and irrigation systems
Mobile phone coverage
Flood trend analysis
Water quality
River embankment erosion
The data sets have different levels of collection and aggregation. The government
works with the SOS and D form for damage and needs assessments. The union secretary or chairman has to fill in the 10 questions of the SOS form within 48 h after the
disaster. The D form has 30 questions, which are filled in and submitted within three
weeks. The system is still largely a paper-based system, whereby forms are manually
summarized at each of the administrative levels, before they are passed on to central
level and digitized. This means that it is often not possible to go back to the data at
ward level. This type of data granularity loss we encountered in more data sources.
The downward arrow in Fig. 18.2 depicts this risk of data granularity loss at each
step up in the government hierarchy. Important data providers are the Department
of Disaster Management of the Government of Bangladesh and the Humanitarian
Coordination Task Team (HCTT), consisting of UN, NGO, and government representatives. For each file, we singled out all the indicators and determined the data type
(excel sheets, relational databases, PDF, text, websites, and geographic information).
(a) Mapping data sets and information products on the information needs
Following the methodology described in Table 18.1, we mapped the 71 information needs on the 15 data sets and information products. We can draw the
following conclusions per cluster level, where for now we do not look at time
constraints. Service locations are not well covered at all with a cumulative coverage of 0.7%.
1 One of the reasons for this might be that information is often
collected by phoning people and asking them to give an overview for their ward
or by conducting a paper-based survey. We did not come across local responders that use an app or GPS to map locations during the floods. Capacity was
also not well covered, varying from relatively easy to monitor capacities such
as the number of boats up to the more difficult to assess coping mechanisms
of affected communities. Damage and needs were covered largely by only two
out of the 15 data sets (JNA and D form). The following data sources match
well the information requirements: JNA (38%), D form (34%), District Disaster
1 Number of times total coverage of service location information needs by data sources divided by
(the number of information needs within service locations) × (the number of data sources).
219
Table 18.2 (continued)
Crisis impact
Operational environment
Flood duration
Security and access
Earlier predictions
News
Time of inundation
Accessibility
Inundated area
Security
Drainage and irrigation systems
Mobile phone coverage
Flood trend analysis
Water quality
River embankment erosion
The data sets have different levels of collection and aggregation. The government
works with the SOS and D form for damage and needs assessments. The union secretary or chairman has to fill in the 10 questions of the SOS form within 48 h after the
disaster. The D form has 30 questions, which are filled in and submitted within three
weeks. The system is still largely a paper-based system, whereby forms are manually
summarized at each of the administrative levels, before they are passed on to central
level and digitized. This means that it is often not possible to go back to the data at
ward level. This type of data granularity loss we encountered in more data sources.
The downward arrow in Fig. 18.2 depicts this risk of data granularity loss at each
step up in the government hierarchy. Important data providers are the Department
of Disaster Management of the Government of Bangladesh and the Humanitarian
Coordination Task Team (HCTT), consisting of UN, NGO, and government representatives. For each file, we singled out all the indicators and determined the data type
(excel sheets, relational databases, PDF, text, websites, and geographic information).
(a) Mapping data sets and information products on the information needs
Following the methodology described in Table 18.1, we mapped the 71 information needs on the 15 data sets and information products. We can draw the
following conclusions per cluster level, where for now we do not look at time
constraints. Service locations are not well covered at all with a cumulative coverage of 0.7%.
1 One of the reasons for this might be that information is often
collected by phoning people and asking them to give an overview for their ward
or by conducting a paper-based survey. We did not come across local responders that use an app or GPS to map locations during the floods. Capacity was
also not well covered, varying from relatively easy to monitor capacities such
as the number of boats up to the more difficult to assess coping mechanisms
of affected communities. Damage and needs were covered largely by only two
out of the 15 data sets (JNA and D form). The following data sources match
well the information requirements: JNA (38%), D form (34%), District Disaster
1 Number of times total coverage of service location information needs by data sources divided by
(the number of information needs within service locations) × (the number of data sources).