1. Overview
South Asia is a flashpoint for natural disasters with profound
societal impacts for the region and globally. Half the worlds
population depends on the regions great rivers, the Indus, Ganges,
and Brahmaputra. The frequent occurrence of floods, combined with
large and rapidly growing populations, ongoing cross-border
conflicts, high levels of poverty, and unstable governments, make
South Asia highly susceptible to humanitarian disasters. The 2007
Brahmaputra floods affecting India and Bangladesh, the 2008
avulsion of the Kosi River in India, and the 2010 flooding of the
Indus River in Pakistan exemplify disasters on scales almost
inconceivable elsewheredisasters devastating local residents and
communities while posing significant threats to the U.S.s and other
nations security interests and assets in the region. The challenges
of mitigating such devastating disasters are exacerbated by limited
flood forecast capability, lack of forecast use and sharing in and
between countries, and the transboundary nature of the hazard.
At the same time, the South Asia situation poses an appropriate
and valuable context for the interdisciplinary study of how
technical and social factors at multiple levels and scales
positively or negatively impact societal vulnerability and
resilience to hazardous events. Despite past flood-related
disasters, high risk of more severe disasters, and the increasing
availability of forecasting, only limited advances have been
achieved in improving forecasting leading to risk mitigation. Many
national meteorological and hydrological agencies in South Asian
countries provide at most a 1-3 day forecast of streamflow and
potential flooding, often with no warning at the upstream border
(examples include Nepal/India and India/Bangladesh). This is in
part due to a lack of streamflow data sharing among the different
countries. Studies undertaken using ensemble weather forecasts have
begun to address technological gaps in meeting specific, regional
flood vulnerability problems (i.e. data sharing, and forecast lead
time) such as for Bangladesh (Hopson & Webster, 2010; Webster
et al., 2010). Consequently, flood prediction partnerships are
suggested as a means to bridge the gap between the existing, global
scale, long lead time weather prediction, and actual implementation
and use of the resulting much-enhanced flood prediction capability
(Webster 2013), since it is clear better flood discharge prediction
will not on its own result in effective outcomes (Syvitiski &
Brakenridge, 2013). There are major gaps in understanding how to
ensure forecasting and monitoring improvements result in enhanced
mitigation of flood risk and flood damage. In such circumstances,
trust is especially important for successful policy implementation
and related behavior changes (e.g., Leach & Sabatier, 2005;
Schlager, 1995; Zafonte & Sabatier, 2004) because individual
decision makers and policy makers are negotiating uncertain
contexts compounded by the delivery of probabilistic forecasts,
which are inherently uncertain.
Our interdisciplinary research will increase understanding of
these interconnected, technical, social, and policy-related
dynamics. It will take an essential step in developing sustainable
approaches to mitigating hazards in policy contexts where new
technologies are being implemented to advance flood-forecasting
science. Specifically, in response to the catastrophic threat posed
by flooding in South Asia and the intellectual challenges they pose
to understand and mitigate them, we have formed the South Asia
Flood Prediction Partnership (SAFPP)an interdisciplinary team of
researchers from the University of Nebraska, U.S. National Center
for Atmospheric Research (NCAR), and the Dartmouth Flood
Observatory at the University of Colorado (DFO). Our partnership
brings together a long history of advancing knowledge in water
policy and behavioral changes; trust studies (including research
related to water resource and hazards); multilevel and latent
variable statistics needed to expand the capabilities of social
science and policy scientists' to simultaneously model multilevel
influences; and geoscience, hydrological modeling, flood
forecasting, and remote sensing. Together the SAFPP proposes linked
activities to mitigate the impact of severe floods in the
Brahmaputra, Ganges, and Indus basins, while advancing a new
framework and methodologies for examining policy implementation and
trust at multiple levels (e.g., micro, meso, and macro levels). Our
work will determine the technical, psychosocial, and policy impacts
of forecast improvements and warning-related societal
communications and responses. Specifically, we hypothesize trust is
a key indicator of policy uptake and successful implementation. We
predict trust will affect how forecast information is used, or
actively resisted, and in turn shape what policy options at
different levels are feasible and considered worthwhile. Our
proposed work will leverage our interdisciplinary teams prior NSF-,
DOE-, NASA-, and USAID-funded research on flood forecasting, near
real time flood remote sensing, measurement and mapping, social
science of trust, and multilevel, structural equation and latent
variable statistics for contextual and multilevel data. As
indicated in our letters of partnership from the Flood Forecasting
Division of the Pakistan Meteorological Department (PMD), the
Regional Integrated Multi-Hazard Early Warning System for Africa
and Asia (RIMES), and the World Bank, these agencies and
organizations will enable access to the decision-making processes,
support, and data we require to conduct our inquiry.
2. Objectives
By accomplishing the following specific objectives we will test
our central hypotheses that: a) innovative data assimilation
approaches will advance big river flood forecasting, b) public
availability of such forecasts will alter societal reaction,
including national and state water policy responses, and c)
successful policy reform, including needed changes in the use of
flood forecasts in decisions affecting tens of thousands of
inhabitants, depends on the presence of specific patterns of trust
at multiple-levels.
Objective 1: Implement long-lead, public-access flood
forecasting systems for the Brahmaputra, Ganges, and Indus basins
and quantify the benefits of data assimilation of satellite-derived
river discharge estimates on improving forecasting skill. The
Climate Forecasting Applications for Bangladesh (CFAB) river flow
forecasting system (Hopson & Webster, 2010), currently
operational only for Bangladesh, will be extended into India and
implemented basin-wide for the Indus. We will then improve existing
forecast modeling through assimilation of satellite microwave
remote sensing of river discharge, creating an enhanced lead-time
(10-15 day) probabilistic river flow forecasting scheme. This can
provide operational updates of the forecast models in-stream flows
and soil moisture conditions. We will test the skill gains provided
by each data assimilation component compared to the initial
forecast system as a function of watershed, spatial-scale, and
forecast lead-time. We will also investigate the predictability
improvements by using multi-model weather forecasting fields in the
hydrologic forecasting system, allowing us to compare and contrast
uncertainty reductions from data assimilation systems to those of
improvements in weather forecasting.
Objective 2: Transform forecast discharge values (flood peak
discharges) into inundation extent maps, as derived from archival
analysis of microwave and optical sensor imagery monitoring actual
inundation extent along the rivers. A unique enhancement to river
flow forecasting will be to transform modeled forecast discharge
peaks into inundation extent maps captured from historic flood
monitoring imagery. Importantly, this provides accurate geolocation
for disaster relief efforts in complicated terrain, where numerical
inundation modeling would normally fail and is in any case
computationally challenging even when data such as channel
bathymetry is available. This will be accomplished using analogue
approaches to select pairings of archived, remote sensing mapping
of inundation extent matched with discharges similar to model
predictions.
Objective 3: Test a model of trust, expand capacity for
contextual/ecological modeling through the use of multilevel and
latent variable statistics, and investigate policy uptake and
implementation as we integrate the forecast information into
reservoir regulation, transboundary water information sharing, and
national and regional disaster planning. We will test our
multilevel model of trust in policy uptake and implementation
contexts, at the same time as we provide in-country education,
capacity building, and technological advancements to ongoing flood
forecasting systems, and investigate ways in which the projects
forecast information and technological development can be
incorporated into ongoing forecasting systems and decision-making
strategies in Bangladesh, India, and Pakistan. To achieve this
objective, we will conduct interviews, focus groups, and surveys
over the same time period as we are providing in-country input and
training workshops for hydrologic engineers and scientists on the
forecasting technologies developed during this project. We will
analyze the data to test the utility and feasibility of competing
statistical methods for testing multilevel social and policy
models.
3. Expected Significance
Intellectual merit. Through this proposal, our South Asia Flood
Prediction Partnership (SAFPP) will significantly advance the
technical and social effectiveness capability of long-lead time
flood forecasting for the Indus, Ganges, and Brahmaputra basins.
The SAFPP will further develop and deploy real-time flood-related
hazards monitoring and forecasting which will, for the first time,
assimilate satellite-based measurements of river discharge and
reservoir storage status into model-based predictions. We also will
assess and reduce potential barriers to policy uptake in the
impacted regions that would interfere with using this vital
forecast information to reduce societal vulnerability.
Specifically, the proposed research is transformative in that it
combines advances in weather forecasting and remote sensing, into a
cohesive framework for successfully integrating and advancing flood
forecasting capacity and subsequent use. Development of data
assimilation approaches to incorporate remotely-sensed discharge
into coupled hydrometeorological forecasting systems can advance
river forecasting, in South Asia, the U.S. and globally. Evaluating
the most beneficial pathways to assimilate remotely-sensed
discharge into these systems will reduce forecast uncertainty.
Doing so places this work at the forefront of research into river
monitoring and forecasting technology development in data-poor
global basins and thus defines future research trajectories in this
field.
At the same time and integrated into the forecasting advances,
we will advance knowledge about trust and its impact on policy
uptake and implementation. Within the social science of trust,
there is a need to develop a comprehensive, multilevel framework to
guide the study of institutional trust and confidence (Li, 2012).
Nowhere is this need greater than in the context of building
resilience to disasters and creating the preconditions for
sustainable development. Thus, in this project, we will advance the
understanding of conceptualization, measurement, and development or
emergence of trust within and across levels of decision making and
within society (Kozlowski & Klein, 2000). Such understanding is
especially important in policy contexts involving multiple actors
facing a potential set of action options through which they may
collectively produce outcomes (e.g., reduce vulnerability to
disaster and increase sustainability) on behalf of others. Related,
our proposed research provides a distinctive opportunity to explore
the conceptual considerations underpinning policy change leading to
sustainability by leveraging our work on trust in institutions. One
of the most pressing considerations facing modern society is how
policy makers confront the myriad sources and forms of uncertainty
which could derail the effectiveness of their preferred
alternatives (Baumgartner & Jones, 2009). Hauser and
Benoit-Barne (2002) suggest issues of trust often arise out of
uncertainty. Consequently, we propose to investigate, a critical
yet under investigated dynamic: How do different levels and
dispersion patterns of different forms of trust impact group
processes and outcomes in contexts of high uncertainty? While there
is a growing literature on the publics trust in government, there
is little research on how trust at different levels impacts
collective decisions involving local and state inter-organizational
collaborations. In an age in which authority is divided among
agencies and across jurisdictions, members of organizations must
engage in collaborative, cross-agency initiatives to fulfill their
own organizations mandate (Michaels, Goucher, & McCarthy,
2006). Thus, we will study the conceptualization, measurement, and
development of trust government officials have in each other and
the collective decision making or advisory groups and processes in
which they participate. Throughout this research, we will advance
the statistical methodology and capabilities of social and policy
scientists to assess and model multilevel and latent variable
influences.
Broader impacts. On a local level, project results will benefit
millions of people exposed to flooding risk in the study region by
significantly improving the warnings (in terms of accuracy, lead
time, accessibility and use) vulnerable residents receive. Indeed,
if a facility such as the one operating at NCAR had been online for
the Indus River in July 2010, making discharge/flood forecasts and
allowing community access, it could have made a tremendous
difference. The Kosi, in 2008 along the India/Nepalese border,
represents another avoidable catastrophe: The flood flow that broke
the levee was not a very large one. The levee could have been
reinforced if the area had the kind of warning that is now possible
and can be readily implemented as proposed in this project. This is
what, in part, this proposal is requesting funding to
accomplish.
We know in order for successful implementation of technologies,
there needs to be acceptance and use. The proposals social science
components are intended to assure successful technical and policy
uptake. We will investigate the extent to which the capacity to
couple discharge forecasts to direct historical observational
imagery provides forecasts of the range of possible inundation
extent in a form accessible to decision makers while still
conveying realistic, probabilistic estimates. This project will
provide all data in a convenient database accessible to
researchers, practitioners and members of the public, and research
results will inform ongoing efforts in the Indus, Brahmaputra, and
Ganges to effectively engage key stakeholders in emergency
management, hazard mitigation and economic development policy and
implementation. But we do not presume that increased forecast skill
and accessibility is enough. We expect the necessity of
understanding the role of trust in the policy making and
implementation process in which they engage to enhance
inter-organizational collaboration in disaster planning. Achieving
the full benefits of scientific and technological breakthroughs
driving the quickly evolving field of flood forecasting requires
the buy-in and cooperation of experts in disaster/emergency
planning and response and robust, regional economic planning.
Policy outcomes of our successful work will include developed
protocols for manipulating reservoir water levels, for example,
intended for rapid execution in circumstances with little
opportunity for reflection. Through participating in this project,
local communities could even be empowered to directly access
relevant flood hazard information.
Importantly, we will work with our partners to assure successful
transfer of the flood forecasting technology to vulnerable
communities in our target sites. We are cognizant of the challenges
in successful technology transfer, both as a matter of technology
as well as politics. We will work closely with our partners who
have extensive expertise in achieving sustainability of technology
transfer. For example, one of our key partners, RIMES, already is
working with communities in Bangladesh on improving access to flood
forecasting warnings. The enhanced model we develop will be
disseminated to other communities with which RIMES presently is
working. Beyond South Asia, our research will be of value to many
communities struggling to make the transition from vulnerability to
resilience and groups engaged in emergency preparedness and
mitigation.
Thus, we will disseminate our research outcomes not only in
South Asia, but also to vulnerable communities elsewhere in Africa,
Asia, and so on, working in partnership with the organizations like
RIMES, CARE-Bangladesh, INIH, FFD, and the World Bank until the
technologies are fully adopted by the various Governments in these
vulnerable nations. We will seek additional funds (e.g., from the
World Bank, USAID, etc.) to support these efforts, as they go
beyond the scope of science and technology development, but they
are critical for long-term impact and sustainability.
The project also provides extensive interaction and knowledge
sharing between U.S. and South Asian researchers and engineers, and
student training in multi-disciplinary, hands-on operational
forecasting systems, data collection techniques, data analysis, and
presentations. Two postdoctoral fellows and GRAs and URAs will
assist with the entire project. We also will request an REU so
additional undergraduates can participate in varied research
activities, such as conducting literature reviews, setting up
surveys, coding, and analysis.
4. Research Plan
Our ambitious research objectives require four years to
complete. Nonetheless, we are planning to achieve substantial
progress and produce tangible accomplishments beginning in the
first year. Arranged by objective, the detailed methods are
provided below..
4.1. Implement long-lead flood forecasting systems.
Climate Forecasting Applications for Bangladesh (CFAB) has been
producing operational flood forecasts for the Ganges and
Brahmaputra Rivers in Bangladesh since 2003, on timescales from
days to months (Hopson & Webster, 2010; Webster et al., 2010).
The proposed methodology builds upon this effort and integrates
adaptable hydrological streamflow multi-model, probabilistic
meteorological/climate forecasts, and satellite and in situ
data.
We will adapt the ensemble forecasting scheme developed by
Hopson and Webster (2010) shown in Figure 1 for the major river
basins in Bangladesh. This will make use of the recently available
THORPEX Interactive Grand Global Ensemble (TIGGE[footnoteRef:1])
multi-center ensemble weather data, with focus on extreme
precipitation[footnoteRef:2] in designing a fully-automated scheme
for 1-15 day predications of river discharge forecasts for the
South Asia region. The hydrological forecast model is a hydrologic
multi-modeling system initialized by NASA and NOAA precipitation
products (e.g., TRMM 3B42, Huffman et al., 2005, 2007; CMORPH,
Joyce et al., 2004; NOAA HydroEstimator) whose states and fluxes
are forecasted forward using TIGGE data products and conditionally
post-processed to produce calibrated probabilistic forecasts of
river discharge for key river reach locations. Already operational
over the Brahmaputra and Ganges river basins, this system will be
extended and calibrated for the Indus basin. [1: See:
http://tigge.ecmwf.int/] [2: See:
http://tparc.mri-jma.go.jp/TIGGE/tigge_extreme _prob.html]
(Figure 2. Collaborative partner RIMES-CEGIS selected pilot
unions of Bangladesh in 2007, selected to receive CFAB long lead
time flood warnings; CFAB 2007 10-day lead-time forecasts (right
panel) showing strong likelihood of severe flooding: ensemble
forecasts - colored lines; observed discharge solid black; danger
level threshold black dashed.)Provide for Public Access. Although
this is the last task represented in Figure 1, providing the public
with truly beneficial flood warning information is a complex
endeavor, and is the subject of much of this proposals research.
However, our team can build from ongoing regional efforts. To
enhance and evaluate the benefits of advanced warnings to
vulnerable Bangladeshi communities living in flood-prone regions,
RIMES, in early collaboration with the Center for Environmental and
Geographic Information Services (CEGIS) and the Forecasting and
Warning Centre (FFWC) of Bangladesh, created a direct dissemination
network to local communities and individuals living in fifteen
pilot unions (Fig. 2, left panel showing original six pilot sites
in 2006) in Bangladesh (CEGIS, 2006). Participants receive flood
hazard warnings based on CFAB model forecasts. The warnings are
disseminated to the pilot unions during the monsoon season, and are
tailored to be understandable to affected communities. For example,
in July and September 2007, severe flooding occurred in the
Brahmaputra basin (Fig. 2, right panel), and warnings were
disseminated to the pilot sites days in advance of the flooding.
This contributed to mitigating the disastrous consequences of the
flood wave on lives and livelihoods. Post flood assessments are
also carried out to assess the effectiveness and drawbacks of the
flood forecast and dissemination system at the community level and
individual level. This pilot dissemination and surveying program is
now being managed and expanded by our collaborative partner, the
Regional Integrated Multi-Hazard Early Warning System for Africa
and Asia (RIMES). These established pilot sites will be locations
in which the University of Nebraska social science team conducts
interviews and focus groups.
Forecast dissemination will also occur at national and
international levels. Since the probabilistic (ensemble) forecasts
can be presented in different ways, during the dissemination and
input workshops, we will work with end-users and prospective end
users to determine the most useful ways to present valuable and
uncertain flood forecast information. The Dartmouth Flood
Observatory (DFO) River Discharge Measurements web site is a model
format on which we can build.[footnoteRef:3] River Watch 2 is an
existing automated processor supported by NASA Earth Science
Research and Applications Programs. In support of our proposed
work, the University of Colorado is making available extensive
space on River Watch 2 for South Asia-focused displays to provide a
prototype portal for both satellite-based present status
information and the model-based discharge prediction and flood
warning information. Thus, Figs. 3 and 4 indicate what has been
entirely missing but is now feasible for South Asia. It provides a
subscene of the present DFO River Watch 2 map view (automated,
updated daily) of current river flow severity status of monitored
sites (Fig. 3), and also a sample time-series of the existing
present and historic status output at one site (Fig. 4). We can
adapt these displays to incorporate forecast and present-status
information (as in Figs. 3 and 4) and create more appropriate map
displays publishable at larger scale for the three river basins of
interest on the DFO and partner organization websites. [3: See:
http://floodobservatory.colorado.edu/CriticalAreas/DischargeAccess.html]
(Figure 5 Daily time series of observed river discharge (solid)
and model nowcast (dash) based on the satellite-based river
discharge estimates for Ganges River at Hardinge Bridge ground
station in Bangladesh. Satellite-derived information at locations
with distance ranging from 63 KM to 1828KM upstream Hardinge Bridge
station were used. From (Hirpa et al., 2013).)Examine the benefits
of data assimilation of satellite-derived river discharge estimates
on improving forecasting skill. Bangladesh is a classic example of
flooding issues exacerbated by international boundaries and lack of
upstream river flow information. Consequently, members of our
research team, in collaboration with the Global Disaster Alert and
Coordination facility in Europe, have used satellite-based daily
measurements of stream widths at multiple locations upstream of
Bangladeshs borders (similar to those shown in Figs. 3 and 4, see
also Fig. 7) to remove this limitation. At certain microwave
wavelengths, there is very little interference from cloud cover
(floods can be measured even when the ground surface is obscured
from optical sensors such as MODIS). Using a strategy first
developed for wide-area optical sensors (Brakenridge, Anderson,
Nghiem, & Chien, 2005), such data can be used to measure river
discharge changes (Brakenridge, Nghiem, Anderson, & Mic, 2007).
As rivers rise and discharge increases, floodplain water surface
area increases. Microwave emission over river measurement sites,
observed from space, can monitor such changes. Our work has: 1)
examined the capability of using these data to track the downstream
propagation of flood waves through India, and 2) evaluated their
use in producing river flow nowcasts (Fig. 5), and forecasts at
1-15 days lead time (Hirpa et al., 2013). The results have
demonstrated the propagation of a flood wave along both river
channels can be tracked reliably in near real time, and therefore
could be incorporated into river flow predictive modeling.
One of the main sources of hydrologic prediction error is due to
uncertainty in the model parameters, as well as routing errors due
to lack of river cross-section information. These problems are
often linked to the lack of reliable ground discharge observation
used for model calibration. This problem can be mitigated by using
discharge estimates derived from satellite-based river measurements
at locations where there are limited ground observations.
Additionally, another large source of hydrologic predictive
uncertainty is lack of knowledge of current in-stream flows and
catchment soil moisture states. Data assimilation can be used to
improve both of these hydrologic uncertainty problems through
optimally combining satellite-based river discharge estimates with
hydrologic model flow states, as well as using the causal link
between in-stream flows and catchment soil moisture. This approach
then modifies a priori state of the model by taking into account
the relative errors in the model simulations and the estimates. In
the past few years, hydrologic data assimilation has become popular
mechanism for reducing forecast uncertainty, in part due to the
availability of a wide range of satellite-based soil moisture and
river information. This proposals research is at the forefront
ofexploring the use of satellite-based river flow measurements in
this context.. Equally relevant for our hydrologic application are
assimilation of in-situ discharge (e.g., Seo et al., 2009;
Moradkhani et al., 2005; Weerts and El Serafi, 2006; Clark at al.,
2008; and Lee at al., 2012), and satellite-based water elevation
(e.g., Montanari et al., 2009).
In this project we propose to improve and extend the lead time
of the near-real time riverflow prediction of the CFAB hydrologic
model using data assimilation of upstream flow information provided
by the remote sensing at several upstream locations. We intend to
use sequential Monte Carlo data assimilation techniques
(Arulampalam et al., 2002) for the assimilation at each model run
time steps. Our recent work demonstrated the value of the remote
sensing data assimilation for operational prediction (Hirpa et al.,
2013) in basins with no ground upstream river discharge
observation, strictly by tracking flood waves. By also assimilating
this information into physical hydrologic models, however, we
anticipate the overall forecast skill of the CFAB scheme could be
improved significantly.
4.2. Transform forecast discharge values (flood peak discharges)
into inundation extent maps.
(Figure 6. The DFO record of flooding in this portion of the
Ganges Basin is shown as light blue (2000 to 20111), light red was
flooding in the 10 days prior to map update date, and red was
current flooding. The numbers indicate River Watch discharge
measurement sites. This is a small subscene of the complete Surface
Water Record display for this region.)NASAs orbital technology has
been used at DFO extensively, since the launch of the twin MODIS
sensors in early 2000 and 2002, to map flooding in South Asia.
Unlike other remote sensing-based organizations active in flood
response, DFO maintains a large and growing archive of such map
data, in digital (GIS) format, and for use in making comprehensive
regional displays indicating the history of inundation as well as
on-going flooding (Fig. 6). The archival flood information is
exceptionally valuable, providing as it does a view of flood
hazard.
This large archive of such mapped inundation resident at DFO
will allow production of an innovative flood prediction product. As
illustrated in Fig. 7, past inundation extent can be matched to the
corresponding remote (Figure 7. Left: MODIS imaging and mapping of
2003 flooding along the Ganges River between river measurement
sites 200 and 201. At site 200 (uncalibrated) peak discharge was
8500 m3/sec. Right: mapping of 2004 flooding. The uncalibrated peak
discharge here is only ~3500 m3/sec. ) sensing-derived discharge
values (the same approach can be used for any ground station sites
for which data output is available publicly). Linkage to the
appropriate inundation map can be provided at the individual site
displays: when a particular discharge and flood threshold is
predicted, the user can call up the inundation that resulted,
historically, from the same values. Mapping inundation maps to the
ensemble of river forecasts produced by the CFAB model could then
produce a range of possible inundation extent scenarios. Such
capability highlights the importance of understanding how and why
end users process this uncertain information. We hypothesize trust
is an important consideration at the individual, group and
intergroup level of interpretive understanding.
4.3. Test a model of trust, investigate policy uptake and
implementation, and expand the capacity for contextual/ecological
modeling through the use of multilevel and latent variable
statistics.
Trust has long been considered a key variable in understanding
cooperation in policy and international relation contexts (e.g.,
Hoffman, 2002), as well as in natural resource management contexts
(e.g., Earle, 2010; Leahy & Anderson, 2008; Poortinga &
Pidgeon, 2006; Stern, 2008; Tennberg, 2007; Winter &
Cvetkovich, 2010). Kingdon (1984), Sabatier and Jenkins-Smith
(1993), and Dye (1995) each propose a conceptualization of policy
making involving the notion of policy communities to probe how the
policy process is shaped by events and actors. Within the policy
community are a core cadre of participants who make most of the
policy domains routine decisions (Baumgartner & Jones, 2009;
Birkland, 1997). These participants function in the policy process
as individuals, members of groups, and members of networks or
systems. In light of this, we hypothesize success in making policy
decisions is based, in part, on patterns of trust that emerge at
individual, group, and systems levels and the interactions between
trust at these levels. Trust likely impacts the diversity of ideas
and the different contributions individuals make, which then
impacts the organizational capacity upon which policy-community
members draw (Innes & Booher, 2003). At the same time,
interpersonal relations are increasingly recognized as contributing
to explaining human behavior in policy contexts (Sabatier &
Weible, 2007). The social component of generating and expanding
knowledge, including trust in institutions, is a function of
collective reasoning (Fleck, 1979): Individuals assume and develop
changed patterns of collective action (Heclo, 1974, p. 306) through
exchanges with others. Potentially, these social practices are
examined continuously and reformed as participants develop insights
into the practices in which they are engaged (Giddens, 1990).
Although much work remains, progress is being made in
conceptualizing, defining, and measuring trust in institutions,
defined most commonly as some combination of positive expectations
and or leading to trust-relevant actions, such as willingness to be
vulnerable to, comply with, and/or otherwise support the
institution (e.g., as reviewed in McEvily & Tortoriello, 2011;
Schoorman, Mayer, & Davis, 2007). The recent emphasis on
advancing the theoretical and methodological foundations of trust
in institutions offered by our interdisciplinary trust research
team (Bornstein, Tomkins, Neeley, Herian, & Hamm, in press;
Hamm et al., in press; Hamm, PytlikZillig, Herian, Tomkins, &
Dietrich, 2012; Hamm et al., 2011; PytlikZillig, Tomkins, Herian,
Hamm, & Abdel-Monem, 2012; Tomkins, PytlikZillig, Herian,
Abdel-Monem, & Hamm, 2010) and by other researchers (e.g.,
Mayer, Davis, & Schoorman, 2006; Nannestad, 2008; Rousseau,
Sitkin, Burt, & Camerer, 1998) has identified critical gaps in
understanding how institutional trust develops and changes over
time. As a key component of our contributions, we posit a
sophistication model of trust development (see Fig. 8) that
hypothesizes as an individual gains experiences with and knowledge
about an institution (i.e., develops sophistication), the
individual will shift from relying on general, trust dispositions,
to more institution-particiularized and then commitment-relevant
trust-constructs (Herian, Hamm, Tomkins, & PytlikZillig,
2012).
(Figure 8. Conceptual diagram of our sophistication model of
trust in institutions)We have been researching our sophistication
model in the contexts of courts, city government, and water
regulatory agencies in the United States. We find the predictive
ability of trust-related constructs has varied across samples (Hamm
et al., in press; Hamm et al., 2011), potentially as a function of
differing levels of sophistication. We are currently examining our
sophistication hypothesis more rigorously in a multi-year
longitudinal study of students who, at initial testing, had very
little experience and knowledge of water regulation institutions.
We have randomly assigned some students to read (over many months)
and learn about water regulation institutions, and others to a
control condition (reading about health-related regulatory
activities). We expect sophistication to change most in the
experimental students, and along with those changes, the bases of
trust in water regulatory institutions to also change.
An important gap in our research and in the field of trust
research at large concerns the potential influence and interaction
of trust constructs as stakeholder sophistication increases and
trust develops and emerges at different ecological or contextual
levels (e.g., micro, meso, macro levels). The vast majority of
trust research, including ours, generally examines trust at a
single level (e.g., the individual, micro, or between-group, meso
level). There is little, if any, research, however, investigating
how trust at different levels might interact or jointly influence
important outcomes (e.g., vulnerability, resilience, policy
outcomes), though it has been argued that a multilevel approach is
critical to understanding how trust impacts policy and other
political behaviors (e.g., Hutchinson & Johnson, 2011). Filling
this gap is important for more fully understanding trust in
institutions, especially given that governance institutions operate
in an inherently multilevel, policy context, in general, and
especially when natural resource decisions cross national borders
(e.g., Hoffman, 2002; Tennberg, 2007).
Moving toward a multilevel model of trust and sophistication.
Examination of trust constructs within a multi-organization and
multilevel context is an ideal place to test, refine, and expand
our sophistication model because amounts and types of knowledge and
experience with an institution (i.e., type of sophistication) are
likely to vary across different levels of the institution and to be
dependent on ones roles and relationships with that institution.
For example, persons involved in the development of flood forecasts
may be quite sophisticated regarding the science and information
provided by the forecasts, but less sophisticated regarding the
politics occurring between stakeholders that might use or be
affected by the use (or non-use) of the forecast information.
Similarly, the public could vary widely in their knowledge,
opinions and appreciation for the complexities involved in decision
making related to a developing forecasting system. Meanwhile,
decision makers may have an added layer of interpersonal and
potentially emotional experience (e.g., frustration, felt
responsibility, discomfort with uncertainty) with the emerging
forecast system. Although trust research has been conducted at
levels ranging from the individual micro to the societal macro
levels (e.g., for macro-level research see La Porta,
Lopez-de-Silanes, Shleifer, & Vishny, 1997; You, 2012),
research on trust constructs and processes that span multiple
levels is rare. Thus, little is known about whether and how trust
at one level (among forecast developers) might affect trust at
another level (decision makers or the public), or the extent to
which emergent processes (Kozlowski & Chao, 2012; Kozlowski
& Klein, 2000) impacts trust at higher-levels.
(Figure 9. Our guiding multilevel trust conceptual model)Our
guiding theoretical model (illustrated by Fig. 9) predicts that
different forms and patterns of trust across multiple contextual
levels will impact intra- and intergroup processes, and thereby
affect the outputs (products) of that performance. At the highest
level, we expect trust in the system as a whole to be impacted via
emergent processes for insiders (e.g., emerging patterns of
relationships with other insiders from within and outside of their
identity groups), and via more direct (e.g., performance outcomes
such as useful and accessible forecasts) processes for the public.
Although the model is comprised of multiple parts, in the present
research we will focus primarily on the central hypothesis that
positive group processes, within and between insider groups (groups
most directly involved in forecast development and use of forecasts
in policy decisions), will be facilitated by optimal (not high or
low) levels of and variability in trust and result in more positive
outcomes and higher levels of overall public trust (likely mediated
by perceptions of performance outcomes). The rationale for this
hypothesis stems from research in the social sciences pertaining to
group functioning and team performance. Within-group or intra-team
trust refers to the levels and patterns of trust (or trust
constructs) that occur among members of the same group. Average
within-group (intra-team) trust refers to the average extent to
which each member of the group trusts each other member (cf.,
intraentity trust, Fang, Palmatier, Scheer, & Li, 2008). Prior
research has found that higher levels of within group trust may
facilitate collaboration (e.g., resulting in high reciprocity and
coordination between group members), but also to less adaptability
to change (e.g., undermining monitoring, promoting features of
group think) (Fang et al., 2008). Thus, we predict that that too
high or low within-group trust will be detrimental to overall group
functioning, especially in contexts that are uncertain, changing,
and unpredictable, such as warning and emergency communication
contexts. Variance in own group trust refers to the overall
variability around the average level of trust in a group to which
someone is a member. We hypothesize that different levels of
individual trust within a group trust may serve different purposes
(with the high-trustors facilitating cooperation and optimism, and
low trustors facilitating monitoring of processes and tasks), and
thus, that there may be an optimal amount of variance of trust in
groups of which one is a part. Without enough variance, different
roles and purposes will not be fulfilled, but too much variance
could cause conflict and undermine group processes. We also will
explore emergent processes and their relationships to variance in
own group trust, such as whether the variance in extent to which
people trust their whole group arises out of variance in dyadic
trust within the group, and/or from differences in dispositional
trust between individual members, and whether one source of
variance is more beneficial than another over time. Between group
trust refers to the levels and patterns of trust that occur between
groups (cf. interorganizational trust, Fang et al., 2008). For
example, between group trust could refer to the trust between two
organizations, two working groups, a working group and a governing
council, a larger group and a sub-group, and so on. Meanwhile,
variance in between group trust refers to the variability in the
extent to which trust exists between groups. Although variables at
different levels (e.g., within vs. between groups) may not
necessarily operate the same, it seems reasonable that, at least in
highly interactive and overlapping groups (such as we propose to
study), too much or too little trust or variance in trust may
impact the extent to which groups effectively collaborate and
monitor one anothers efforts.
Study 1. Initial insider measure development and context
description. The objective of Study 1, which will be conducted
during months 1-6 of the project, is to develop measures and
establish a foundational understanding of the context in which our
model testing will take place. This study will be part 1 of an
exploratory, sequential design (Creswell & Plano Clark, 2007;
Small, 2011), and will be primarily qualitative in nature. Data
sources will include structured notes from focus groups, and
interviews (which will also be audiotaped in case of the need for
clarification). Participants in the focus groups and interviews
will include engineers and scientists currently producing weather
and flood forecasts, those directly involved in the ongoing efforts
to put the sophisticated flood forecasts to use in planning for
disasters and regional economic development, including World Bank
planners, NGO specialists working in the field on mitigation
strategies, government policy makers, and residents.
Our procedures will involve traveling to the study areas early
in Year 1, and conducting approximately 3 focus groups (one per
country), and up to 15 semi-structured interviews of different
stakeholders (5 per country) in order to sample a range of
participants from each country and watershed, to explore the
elements of our model (Fig. 9) and initial drafts of our measures
of model elements. We will recruit relevant stakeholders for
interviews and focus groups with assistance from CARE-Bangladesh,
INIH, FFD, World Bank, and RIMES. Initial drafts of measures of the
trust and process variables will be adapted from prior work (e.g.,
De Jong & Dirks, 2012; Fang et al., 2008; Hamm et al., 2011),
while measures of the products (individual and team performance,
performance outcomes, such as related to the accessibility and
utility of the forecasts) and system sustainability will be
developed as part of this research, with input from stakeholders.
We expect interviews to last about 1 hour, and focus groups to last
2 hours. During the workshops, focus groups and interviews, in the
tradition of Glaser and Strausss (1967) grounded theory, we will
use generative questions (Trochim, 2005) to discuss each of the
model components (intra and intergroup trust, group processes,
products/performance, and system sustainability) and subcomponents
(e.g., the subcomponents of trust, the importance of expertise and
competence assessments, the impact of uncertainty), as they relate
to key aspects of the forecasts and events related to flood
forecasts (e.g., actual flooding). The in-depth, semi-structured
nature of our protocols will give respondents the opportunity to
provide detailed explanation and clarification (Lewis, 2003), and
allow us to probe a consistent set of issues while hearing a range
of perspectives (Berg, 1998; Hughes, 2002), thereby revealing
strengths and weaknesses in our emerging theoretical model in this
specific context. Note that the strategy of developing new
variables dependent on what we learn from the focus groups and
interviews, as well as our observations of dissemination workshops,
makes protocol development an ongoing feature of the research
rather than an activity we complete at the outset. Study 1 will
begin this process, which will be continued in Studies 3 and 4.
During the interviews we will also use cognitive interview
techniques (Davison, Vogel, & Coffman, 1997) to better
understand participant thoughts and reactions to the survey
measures we develop, and for validation and fine tuning. A trained
researcher will be present at each focus group to take notes. S/he
also will listen to and take structured notes from the audio
recordings of the interviews, and these notes, along with the notes
from the flood forecast use dissemination workshops, will form the
qualitative data base.
Analyses of data from the Study 1 focus group and interview
notes will be conducted throughout Year 1, using Atlas.ti or other
qualitative coding software. Two coders will use a draft coding
protocol to analyze a subset of the notes for themes relating to
the model components as well as additional themes that may become
apparent during analyses. Coders will then meet with other research
team members to expand, revise, and refine the codes as needed.
Each document in the corpus will be coded by at least one coder,
with a subset of the documents coded by two coders to estimate
intercoder reliability.
Study 2. Testing and refinement of insider measures. Study 2
will be conducted at the end of Year 1 of the project (at about the
time of the dissemination workshops) and is part 2 of the
exploratory sequential design. The objective will be to gather data
pertaining to the statistical/psychometric and construct validity
of the measures (developed as part of Study 1) and to provide
baseline data for Study 4 (Specific Aim 5). Participants will
include all participants in the 3-day dissemination workshops (Type
A participants, n 30, about 10 persons per country) which will
include engineers and scientists directly involved in the ongoing
efforts to put the sophisticated flood forecasts to use in planning
for disasters and regional economic development; and other members
of involved institutions who are both knowledgeable and not about
the use of sophisticated flood forecasting but not directly
involved in the workshop activities (Type B participants, n 120,
approximately 40 per country), including RIMES experts, World Bank
specialists, scientists and engineers currently producing weather
and flood forecasts but not attending the workshops, and social
scientists and NGO specialists working in the field on mitigation
strategies. Participants will be recruited using multiple
strategies with the assistance of our in-country collaborators
(i.e. CARE-Bangladesh, etc.) and, if relevant, snowball sampling
where those more centrally involved invite colleagues who are
interested and have varied levels of knowledge about the use of
sophisticated flood forecasting in developing disaster preparedness
and regional economic plans).
To ensure efficient and appropriate measurement, surveys will be
customized depending on the participants role or relationship to
using sophisticated flood forecasting in developing disaster
preparedness and regional economic plans (e.g., specific questions
for workshop participants or relevant group members will differ
from questions for more marginally involved participants). The
survey will include measures of multilevel trust, process
perceptions, and perceptions of team performance outcomes and
sustainability of the system as whole and designed to take
approximately 10-15 minutes. Initial quantitative analyses will
include item analyses, scale reliabilities, exploratory factor
analyses, correlation, and multiple regression analyses designed to
examine both scale reliability/validity and hypothesized
relationships between variables. Power analyses indicate that
approximately 150 participants will be sufficient in our model to
find effects as small as .22 at a power level of .8 and p < .05
(Soper, 2013). In addition, these data will be used in the more
sophisticated investigations in Study 6.
Study 3. Development of public measures. During Years 1 and 2,
measures of public trust in the flood forecast and warning systems
will also be developed based on our prior measures. Participants in
this study will be students from area universities who will serve
as proxies for the general public (N 300 total, 100 per country).
Procedures will be primarily survey-based. After a draft of the
survey is developed and translated into the appropriate languages,
at the end of Year 1 (around the time of the first workshops) we
will conduct 3 cognitive interviews per country with students to
ensure the understandability and appropriateness of the questions.
The students will be recruited through our in-country collaborators
and chosen for their interest in flood and forecasting topics.
Because they may not know much about our teams efforts, we will
prepare background information about the ongoing activities and
flood forecasting in general for students to read before they
complete the survey (which will assess trust and perceptions of
team performance outcomes and of the system as whole, and will be
designed to take about 10-15 minutes) and then engage with a
researcher in a cognitive interview for approximately 30-45
minutes. They will be paid a $25 incentive for their time and
effort. After refining the questions and background materials based
on the student interviews, the reading and questions will finalized
into both a paper and web-based survey. During Year 2, additional
students (N 300; 100 per country) will be recruited from courses
related to weather, climate and forecasting. These students will
read background material, varying in detail and depth, about our
project prior to completing the measures. The varied background
given to the students, as well as natural variation in student
knowledge and experience with the topics, will simulate differences
in sophistication that we expect to find in the general public. We
will conduct confirmatory factor analyses on the items to examine
their dimensionality and to determine replication of prior
findings. Initial analyses will assess scale reliability, using
model-based reliability estimates (omega), and construct
relationships, using structural equation modeling. Power analyses
indicate that 300 participants will be more than sufficient to
detect effects as small as .15 in our model (Soper, 2013). In
addition, these data will be used in the more sophisticated
investigations in Study 6.
Study 4. Simultaneous, explanatory, mixed-method, longitudinal
study. This study uses an explanatory, mixed-method design in which
qualitative study of the model conceptually outlined in Figure 1 is
designed to help explain and understand (as well as triangulate)
quantitative data pertaining to the model (Small, 2011). In
addition, the study will continue the work started in Study 1 to
construct our ultimate outcome variables (e.g., final assessments
of perceptions of key events that occur during the project time
period). This study will examine our previously-described optimal
trust working hypothesis. Participants will include the same target
groups (i.e., Types A and B) as described for Study 2. Our
procedures will include the use of surveys developed during Studies
1 and 2 administered at the end of Year 1 (as part of Study 2), and
again at the end or Years 2 and 3 (in conjunction with the Year 3
workshops, which will include some of the same participants as
involved at the end of Year 1). Initial analyses of surveys will
include longitudinal multilevel modeling procedures that take into
account the highly nested nature of the data (time points nested
within individuals nested within groups and organizations); but
once again, these data will also be investigated in Study 6.
To obtain the explanatory qualitative data we will conduct
twelve focus groups (one with each Type A/B participants in each
country at the end of Year 1 (six focus groups) and likewise at the
end of Year 3 (six more groups), with approximately 8-10
participants per group) as well as approximately 15 interviews (5
per country) conducted and spaced out between the two sets of focus
groups. Focus groups will be conducted as part of the workshops;
interviews will be conducted using distance methods (e.g., with the
assistance of an in-country partner trained to conduct the
interview, and a research co-interviewer in attendance via Skype or
conference call). For the focus groups and interviews (which also
will be audiotaped), we will use semi-structured protocols to
examine how decision process outcomes result from initial
conditions. This will provide important context for understanding
changes in trust in the process of development and use of the
forecasts. As was the case in Study 1, a trained research assistant
will use a structured note taking protocol to gather data while
attending the focus groups and from the audio tapes of the focus
groups and interviews. The coding protocol for the qualitative data
will build upon findings from Study 1 but also will be adapted as
needed to take into account events taking place during the
study.
Study 5. Public survey study. A large-scale survey will
investigate public perceptions of and trust in flood forecast
systems, science, and forecast information during Years 2 and 4 of
the study. This survey, initially developed as part of Study 3,
will take approximately 10-15 minutes. Participants will be a
sample of the general public (100 per country, across 3 sites per
country or 9 sites total) who will be asked their opinions about
utility of flood forecasting in mitigating devastation from floods
(as in Study 3, N = 300+ will allow us to achieve the power we need
for our initial analyses). Survey procedures will be used, and
measures will include perceptions of team performance outcomes and
of the system as whole. Survey participants will have the
opportunity to (but will not be required to) read or hear
background information about the project previously developed for
and tested with students. Initial analyses of the survey data will
aim to replicate those used with Study 3 data, but now with data
from the general public. Once again, however, we expect to also use
these data in our Study 6, discussed next.
Study 6. Evaluation of methods for statistical modeling of
multilevel trust. Multilevel research efforts are often hindered by
an inability to simultaneously and statistically consider multiple
levels of ecological and/or environmental influence (Bovaird,
2007). Ecological models such as the posed model of trust are often
theoretically unified, but tested in a piecemeal fashion.
Innovative modeling methods that overcome these limitations and are
applicable to studying such multilevel determinants are needed.
Multilevel structural equation modeling (MSEM) presents a
statistical modeling framework that synthesizes multilevel modeling
(MLM) appropriate for considering the effects of complex sampling,
with broader techniques appropriate for integrating latent
variables and measurement assumptions into multivariate linear
models, thus enabling simultaneous evaluation of comprehensive
ecological or contextual systems (Bovaird, 2007). The goals of this
study are to: 1) continue to expand the capacity of current MLM and
SEM frameworks to allow designation of complex levels of influence;
2) determine the sufficient data conditions necessary for
implementing both complex univariate MLM models and complex MSEM;
and 3) evaluate recently developed commercial and research software
to enable implementation of the methodologies evaluated in this
study. The first goal of this study will be achieved by
implementing a statistical modeling framework that allows for
consideration of more than two levels of an ecological MSEM for the
multilevel theory of trust developed and evaluated in Studies 1-5.
The data obtained in Studies 1-5 will be integrated to provide
substantively valid estimates of population parameters based on
relationships posed in the model of multilevel trust. In pursuit of
the second goal, Monte Carlo simulation methods will be used to
validate the statistical modeling framework with simulated data.
Based on parameter estimates from empirical data collected in
studies 1-5, Monte Carlo simulation methods will then be used to
determine the data conditions necessary to confidently apply
substantively relevant MLM and MSEM models. Finally, the third goal
will be achieved by evaluating accessible software for implementing
the expanded MSEM in the Mplus, SAS, WinBUGS, and R computing
environments. This study will provide a significant expansion of
the MSEM framework to allow consideration of complex realistic
multilevel ecological models. The methods developed and evaluated
in this study will be applicable to all research areas in social
sciences that consider ecological or contextual models of behavior
or development. This work is essential for answering the call for
multilevel rather than single level approaches to understanding how
trust impacts policy and other political behaviors (Hutchinson
& Johnson, 2011).
Both the empirical and simulation phases of this study will
compare model estimates from the following softwares: the recently
released Mplus version 7 (Muthn & Muthn, 1998-2012) which now
allows complex hierarchical nesting; a two-step approach using a
combination of Mplus version 7 and procedures/packages from the SAS
9.3 and R software environments; and WinBUGS for implementing
Bayesian methods through Markov Chain Monte Carlo (MCMC). Of
particular interest is the newly-released Mplus version 7 which has
the expanded ability to model three-or-more nested hierarchical
levels or cross-classified non-nested levels with multivariate
latent variable models. Mplus 7 utilizes a Bayesian approach
through MCMC, as does WinBUGS, to implement such complex multilevel
models; however, the software has only been available for a limited
time (as of November 2012) and has not been thoroughly evaluated.
Prior to the release of this version of Mplus, there were limited
options (i.e. only two levels) for estimating complex multivariate
latent variable model such as required in simultaneous estimation
of complex multilevel ecological models. Thus, the proposed work is
timely and has the potential for broad impacts across the social
and behavioral sciences.
Also of particular focus in this studyparticularly goal 2 are
the methodological limitations posed by multilevel ecological
research pertaining to sample size and geographical nesting.
Generally speaking, a large sample size (often j = 50-100 macro
level units) is required for multilevel modeling and particularly
MSEM primarily because of the model complexity and estimation
methods used to obtain parameter estimates. Obtaining a large
sample size at the micro/individual level is often not a problem,
but obtaining a sufficiently large sample size at the macro level
is. In the current study, the micro-level population is relatively
large and accessible, while the macro-level populations (regions
and countries) are finite and small. Song and Lee (2004) have shown
that Bayesian methods similar to those implemented in Mplus and
WinBUGS can be particularly effective for estimating structural
models with small samples. In addition, geographically-based
macro-level units vary greatly in population size, making it
difficult if not impossible to have a balanced design.
5. Relation to Results of Prior NSF Research, Other Research,
& Long-Term Goals
The proposed research brings the Colorado teams (NCAR and DFO)
prior research on flood forecasting together with the Nebraska
teams prior NSF-funded research on trust and stakeholder
engagement. In doing so we are able to advance the fields of flood
forecasting and trust in the context of technical and policy
acceptance and implementation. The inclusion of the sophisticated
statistical modeling and latent variable analysis components of
this project address the challenge of assessing trust and policy
uptake at different scales. Thus, as a team, we will work towards
our long-term, shared goals of reducing flood vulnerability through
advancing flood forecasting and its use by agencies and groups, on
behalf of vulnerable individuals.
We will leverage our successful NASA-funded research feasibility
project designed to better define the pathway to be taken to full
and sustainable implementation of a flood mapping processor, a
merger between an automated, near real time (NRT), MODIS
sensor-based flood map product and a complementary, radar
frequency, Envisat ASARbased global flood mapping processor
(Kleuskens, et al., 2011; Westerhoff et al., 2010). Also, in
collaboration with end users, the research team is implementing, on
a trial basis, coupled flood discharge and inundation forecasting
for Bangladesh. During our dissemination and input workshops with
end users we will determine, for example, which forecast locations
would likely provide the earliest demonstration of specific
enhancements to end-user capabilities and operations. This will
contribute to increasing local capacity to plan for the flood
threat. Building on the advances in flood forecasting by NCAR and
DFO, we are developing a valuable product to provide decision
makers, within and outside of government, with maps indicating
anticipated extent of flood plain inundation. Such work and our
comparative examination of trust in multiple organizations and
among individuals integral to the successful use of flood forecasts
is made possible by leveraging our exceptional access to and
positive relationships with them made possible through our
relationships with CARE-Bangladesh, INIH, PMD, World Bank, and the
World Food Programme.
As key parts of each of our NSF-funded projects relating to
trust (Testing a Three-Stage Model of Institutional Confidence
across Branches of Government (SES-1061635); Law and Social Science
Postdoctoral Fellowship: Trust and Confidence (SES-1228559);
Understanding the Role of Trust in Cooperating with Natural
Resource Institutions (SES-1154855); SBES: Medium: Investigating
the Role of Distrust in Unauthorized Online Activities Using an
Integrated Sociotechnical Approach; IGERT: Resilience and Adaptive
Governance in Stressed Watersheds (DGE-0903469)), we have been
developing reliable and valid measures of trust-related constructs
and testing and refining an interdisciplinary theory of the
development of institutional trust. In the cybersecurity project,
we are examining the relevance of the construct of distrust and its
impact on hacking behaviors. The theory-testing and measures
development of the trust measures is currently ongoing in all the
projects. Relevant publications include: Bornstein et al. (in
press); Hamm et al. (in press); Hamm et al. (2011); PytlikZillig et
al. (2012); Tomkins et al. (2010). Relevant papers include: Hamm et
al. (2012); Tomkins, PytlikZillig, Herian, and Hoppe (2012). The
proposed project will build on this prior work, allowing us to
assess and refine our sophistication model in a dynamic,
real-world, societally-important, policy context. It also provides
us with the opportunity to go beyond the individual or micro level
and test the potential influence and interaction of trust
constructs at the meso and macro levels. At the same time, the
proposed research allows us to explore further the role of time, as
we study trust constructs as they evolve, thus providing us another
opportunity to assess sophisticationnot simply as a static,
point-in-time construct measured cross-sectionallybut also as a
longitudinal and developmental phenomenon.
The present work also builds on our NSF-funded work in public
engagement (Developing a Social-Cognitive, Multilevel,
Empirically-Based Model of Public Engagement for the Shaping of
Science and Innovation Policy (SBE-0965465)) which uses U.S.
students as research participants and experimentally investigates
methods for enhancing public engagement and the publics use of
complex science-related information in their policy recommendations
concerning nanotechnology development and regulation. This research
has included examination of trust in science and scientists and has
involved about 1,000 science majors in considering ethical, legal
and social issues related to nanotechnology (PytlikZillig &
Tomkins, 2011; PytlikZillig et al., 2011; PytlikZillig et al.,
2013). The current proposal will allow us to build on our
experiemental work in real-world, international contexts. We have
also conducted NSF-Funded work in water contexts (Knowledge
Discovery and Information Fusion Tools for Collaborative Systems to
Adaptively Manage Uncertain Hydrological Resources (IIS-0535255).
This project studied complex hydrologic data and decision making in
the U.S., and provided the basis for studying water resource
institutions, along with considering aspects of trust and social
justice in this policy arena (Bornstein et al. (2009)). The
proposed research takes that research into international
contexts.
Finally, the proposed project will build on the Department of
Education-funded and related research that has advanced statistical
research on multilevel modeling (Bovaird, 2007; Bovaird &
Embretson, 2008; Bovaird & Koziol, 2012; Locker, Hoffman, &
Bovaird, 2007) and latent variables (Little, Bovaird, &
Widaman, 2006) as well as modeling contextual effects in
longitudinal studies (Little, Bovaird, & Card, 2007).
REFERENCES
Adler, R.F. et al, 2003: The Version-2 Global Precipitation
Climatology Project (GPCP) Monthly Precipitation Analysis
(1979-Present), J. of Hydromet., 4 (6).
Arulampalam, M.S., S., Maskell, N., Gordon, and T., Clapp, 2002:
A tutorial on particle filters for online nonlinear/non-Gaussian
Bayesian Tracking. IEEE Tran. Signal Proc. 50, 174-88.
Aubert, D., C. Loumagne, and L. Oudin, 2003: Sequential
assimilation of soil moisture and streamflow data in a conceptual
rainfall-runoff model. J. Hydrol. 280, 145-161
Brakenridge, G. R., S. V. Nghiem, E. Anderson, and R. Mic
(2007), Orbital microwave measurement of river discharge and ice
status, Water Resources Research, 43(W04405,
doi:10.1029/2006WR005238).
Brakenridge, G., R., Cohen, S., Kettner, A.J., De Groeve, T.,
Nghiem, S.V., Syvitski, J.P.M., Fekete, B.M., Calibration of
satellite measurements of river discharge using a global hydrology
model, J. Hydrol. (2012), doi:
http://dx.doi.org/10.1016/j.jhydrol.2012.09.035.
Brocca, L., F. Melone, T. Moramarco, W. Wagner, V. Naeimi, Z.
Bartalis, and S. Hasenauer, 2010: Improving runoff prediction
through the assimilation of the ASCAT soil moisture product.
Hydrol. Earth. Sys.. Sci. 14, 1881-1893,
doi:10.5194/hess-14-1881-2010
Brocca, L., T. Moramarco, F. Melone, W. Wagner, S. Hasenauer,
and S. Hahn, 2012: Assimilation of surface and root zone ASCAT soil
moisture products into rainfall-runoff modeling. IEEE T. Geosci.
Remote Sens. 50, 2534-2555. Doi:10.1109/TGRS.2011.2177468
CEGIS (2006), Sustainable end-to-end climate/flood forecast
application through pilot projects showing measurable improvements,
CEGIS Base Line Rep, 78.
Clark, M.P., D.E., Rupp, R.A., Woods, X., Zheng, R.P., Ibbitt,
A.G., Slater, J. Schmidt and M.J, Uddstrom, 2008: Hydrological data
assimilation with the ensemble Kalman filter: Use of streamflow
observations to update states in a distributed hydrological model.
Adv. Water Resour. 31, 1309-24.
Cohen, S., A. Kettner, J. Syvitski, and F. B.M. (2011), WBMsed:
a distributed global-scale daily riverine sediment flux model -
model description and validation, Computers & Geosciences, In
Review.
Crow, W.T., and D. Ryu, 2009: A new data assimilation approach
for improving runoff prediction using remotely-sensed soil moisture
retrievals. Hydrol. Earth. Sys.. Sci. 13, 1-16
doi:10.5194/hess-13-1-2009
De Groeve, T., and P. Riva (2009), Early flood detection and
mapping for humanitarian response, Proceedings of the 6th
International ISCRAM Conference Gothenburg, Sweden, May 2009, J.
Landgren and S. Jul, eds.
De Groeve, T., Z. Kugler, and G. R. Brakenridge (2006), Near
Real Time Flood Alerting for the Global Disaster Alert and
Coordination System, Proceedings ISCRAM2007 (B. Van de Walle, P.
Burghardt and C.Nieuwenhuis, eds.), 33-39.
Draper, C., J.F. Mahfouf, J.C. Calvet, E. Martin, and W. Wagner,
2011: Assimilation of ASCAT near-surface soil moisture into the SIM
hydrological model over France. Hydrol. Earth Syst. Sci. 15,
3829-3841. doi:10.5194/hess-15-3829-2011
Hain, C.R., W.T. Crow, J.R. Mecikalski, M.C. Anderson, and T.
Holmes, 2011:An intercomparison of available soil moisture
estimates from thermal infrared and passive microwave remote
sensing and land surface modeling. J. Geophys. Res. Let. 116,
D15107. doi:10.1029/2011JD015633
Hirpa, F.A, T. Hopson,T. De Groeve, G.R. Brakenridge, M.
Gebremichael and P. J. Restrepo, 2012: Upstream Satellite Remote
Sensing for River Discharge Prediction: Application to Major Rivers
in South Asia. Remote Sen. of the Environt.
DOI:10.1016/j.rse.2012.11.013
Hopson, T.M. and P.J. Webster, 2010: A 1-10-Day Ensemble
Forecasting Scheme for the Major River Basins of Bangladesh:
Forecasting Severe Floods of 2003-07, J. of Hydro., 11,
618-641.
Huffman, G. J., R. F. Adler, S. Curtis, D. T. Bolvin, and E.
J.Nelkin (2005). Global rainfall analyses at monthly and 3-hr time
scales. Measuring Precipitation from Space: EURAINSAT and the
Future, V. Levizzani, P. Bauer, and J. F. Turk, Eds., Springer, 722
pp.
Huffman, G. J., Bolvin, D. T., Nelkin, E. J., Wolff, D. B.,
Adler, R. F., Gu, G., . . . Stocker, E. F. (2007). The TRMM
multisatellite precipitation analysis (TMPA): Quasi-global,
multiyear, combined-sensor precipitation estimates at fine scales.
Journal of Hydrometeorology, 8(1), 38-55.
Kleuskens, M., R. S. Westerhoff, and J. Huizinga (2011),
Operational Flood Mapping: A Pilot Study in the Mekong Area, paper
presented at Proceedings of the International Symposium of Remote
Sensing of the Environment, Sydney, Australia, April 10-15
2011.
Lee, H.S., D.J. Seo, Y. Lui, V. Koren, P. McKee, and R. Corby,
2012: Variational assimilation of streamflow into operational
distributed hydrologic models: effect of spatiotemporal scale
adjustment. Hydrol. Earth Syst. Sci. 16, 2233-2251.
doi:10.5194/hess-16-2233-2012
Montanari, M., R. Hostache, P. Matgen, G. Schumann, L. Pfister,
and L. Hoffman, 2009: Calibration and sequential updating of a
coupled hydrologic-hydraulic model using remote sensing-derived
water stages. Hydrol. Earth Sys. Sci. 13, 367-380.
Doi:10.5194/hess-13-367-2009
Moradkhani, H., K. Hsu, H.V. Gupta, and S. Sorooshian, 2005:
Uncertainty assessment of hydrologic model states and parameters:
sequential data assimilation using particle filter. Water Resour.
Resear. 41, W05012, doi:10.1029/2004WR003604.
Parajka, J., V. Naeimi, G. Bloschl, W. Wagner, R. Merz, and K.
Scipal, 2006: Assimilating scatterometer soil moisture data into
conceptual hydrologic models at the regional scale. Hydrol. Earth
Sys. Sci. 10, 353-368. Doi:10.5194/hess-10-353-2006
Pauwels, V.R., R. Hoebeb, N.E. Verhoest, and F.P. De Troch,
2001: The importance of the spatial patterns of the remotely sensed
soil moisture in the improvement of discharge prediction for small
scale basins through data assimilation. J. Hydrol. 251, 88-102
Reichle, R.H, W.T. Crow, C.L. Keppenne, 2008: An adaptive
ensemble Kalman filter for soil moisture data assimilation. Water
Resour. Resear., 44, W03423, doi:10.1029/2007WR006357.
Seo, D.J., L., Cajina, R. Corby, and T. Howieson, 2009:
Automatic State updating for operational streamflow forecasting via
variational data assimilation. J. Hydrol. 367, 255-275.
Syvitski, J. P. M., and G. R. Brakenridge (2013), Causation and
avoidance of catastrophic flooding along the Indus River, Pakistan,
GSA Today, 23 (1), doi: 10.1130/GSATG1165A.1131.
Verdin, J., K. Verdin, H. Gadain, and M. Mathis (2011),
Development and Implementation of Globally Applicable Methods for
Characterization of Flood Hazards: Project Completion Report, in
Project performed by the U.S. Geological Survey, in partnership
with UNEP, and with the funding support of the USAID Office of U.S.
Foreign Disaster Assistance (OFDA). edited, p. 28.
Webster, P. J., and e. a. T. M. Hopson (2004), Regional
application of Monsoon Dynamics: Implementation of a three-tier
flood and precipitation forecasting scheme for Bangladesh and
surrounding regions, CLIVAR Contribution to Exchanges, 30.
Webster, P.J., J. Jian, T.M. Hopson, C.D. Hoyos, P. Agudelo,
H-R. Chang, J.A. Curry, R.L. Grossman, A.R. Subbiah, 2010:
Extended-Range Probabilistic Forecasts of Ganges and Brahmaputra
Floods in Bangladesh, Bulletin of the American Meteorological
Society, 91 (11), 1493-1514.
Webster, P. J., V. E., Toma and H. M. Kim, 2011: Were the 2010
Pakistan floods predictable? Geophys. Res. Lett. 38, L04806.
Webster, P. J., 2013: Improve weather forecasts for the
developing world. Nature 493, 17-19.
Weerts, A.H., and G.Y.H., El Serafy, 2006: Particle filtering
and ensemble Kalman filtering for state updating with hydrological
conceptual rainfall-runoff models. Water Resour. Resear. 42,
W09403, doi:10.1029/2005WR004093.
Westerhoff, R., J. Huizinga, M. Kleuskens, R. Burren, and S.
Casey (2010), Operational Satellitebased Flood Mapping using the
Delft-FEWS System, paper presented at Proceedings of the ESA Living
Planet Symposium, Bergen, Norway, June 28-July 2, 2010.
Baumgartner, F. R., & Jones, B. D. (2009). Agendas and
instability in American politics (2nd ed.). Chicago: University of
Chicago Press.
Berg, B. L. (Ed.). (1998). Qualitative research methods for
social science (3rd ed.). Toronto, Ontario: Allyn and Bacon.
Birkland, T. A. (1997). After disaster: Agenda setting, public
policy, and focusing events. Washington, DC: Georgetown University
Press.
Bornstein, B. H., Tomkins, A. J., Michaels, S., Samal, A., Nam,
Y., Zellmer, S., . . . Olson, D. (2009). Social justice and water
sustainability and management. In A. Kakanowski & M. Narusevich
(Eds.), Handbook of social justice (pp. 271-282). Hauppauge, NY:
Nova Science.
Bornstein, B. H., Tomkins, A. J., Neeley, E. M., Herian, M. N.,
& Hamm, J. A. (in press). Reducing courts failure-to-appear
rate by written reminders. Psychology, Public Policy, and the Law,
18. doi: 10.1037/a0026293
Brakenridge, G. R., Anderson, E., Nghiem, S. V., & Chien, S.
(2005). Space-based measurement of river runoff. EOS, Transactions
of the American Geophysical Union, 86.
Brakenridge, G. R., Nghiem, S. V., Anderson, E., & Mic, R.
(2007). Orbital microwave measurement of river discharge and ice
status. Water Resources Research, 43(W04405,
doi:10.1029/2006WR005238).
Creswell, J. W., & Plano Clark, V. L. (2007). Mixed methods
research. London: Sage.
Davison, G. C., Vogel, R. S., & Coffman, S. G. (1997).
Think-aloud approaches to cognitive assessment and the articulated
thoughts in simulated situations paradigm. Journal of Consulting
and Clinical Psychology, 65(6), 950-958.
De Jong, B. A., & Dirks, K. T. (2012). Beyond shared
perceptions of trust and monitoring in teams: Implications of
asymmetry and dissensus. Journal of Applied Psychology, 97(2),
391-406.
Dye, T. (1995). Who is running America? The Clinton years.
Englewood Cliffs, NJ: Prentice Hall.
Fang, E., Palmatier, R. W., Scheer, L. K., & Li, N. (2008).
Trust at different organizational levels. Journal of Marketing,
72(2), 80-98.
Fleck, L. (1979). Genesis and development of a scientific fact.
Chicago: University of Chicago.
Giddens, A. (1990). The consequences of modernity. Cambridge,
MA: Polity Press.
Glaser, B. G., & Strauss, A. (1967). The discovery of
grounded theory: Strategies for qualitative research. Chicago,
Illinois: Aldine De Gruyther.
Hamm, J. A., PytlikZillig, L., Herian, M. N., Bornstein, B. H.,
Tomkins, A. J., & Hoffman, L. (in press). Deconstructing public
confidence in state courts. Journal of Trust Research.
Hamm, J. A., PytlikZillig, L. M., Herian, M. N., Tomkins, A. J.,
& Dietrich, H. (2012). Trust and intention to comply with a
water allocation decision: The moderating roles of knowledge and
consistency. Manuscript submitted for publication.
Hamm, J. A., PytlikZillig, L. M., Tomkins, A. J., Herian, M. N.,
Bornstein, B. H., & Neeley, E. (2011). Exploring separable
components of institutional confidence. Behavioral Sciences and the
Law, 29, 95-115.
Hauser, G. A., & Benoit-Barne, C. (2002). Reflections in
rhetoric, deliberative democracy, civil society, and trust.
Rhetoric & Public Affairs, 5(2), 261-275.
Heclo, H. (1974). Modern social politics in Britain and Sweden:
From relief to income maintenance. New Haven, Connecticut: Yale
University Press.
Herian, M. N., Hamm, J. A., Tomkins, A. J., & PytlikZillig,
L. M. (2012). Public participation, procedural fairness, and
evaluations of local governance: The moderating role of
uncertainty. Journal of Public Administration Research and Theory,
22, 815-840.
Hughes, M. (2002). Interviewing. In T. Greenfield (Ed.),
Research methods for postgraduates (2nd ed., pp. 209-217 ). New
York: Oxford University Press.
Innes, J. E., & Booher, D. E. (2003). Collaborative
policymaking: Governance through dialogue. In M. A. Hajer & H.
Wagennar (Eds.), Deliberative policy analysis: Understanding
governance in the network society (pp. 33-59). Cambridge, UK:
Cambridge University Press.
Kingdon, J. W. (1984). Agendas, alternatives, and public
policies. Boston: Little Brown.
Kozlowski, S. W. J., & Chao, G. T. (2012). The dynamics of
emergence: cognition and cohesion in work teams. Managerial and
Decision Economics, 33, 335354.
Kozlowski, S. W. J., & Klein, K. J. (2000). A multilevel
approach to theory and research in organizations: contextual,
temporal, and emergent processes. In K. J. Klein & S. W. J.
Kozlowski (Eds.), Multilevel theory, research and methods in
organizations: foundations, extensions, and new directions (pp.
3-90). San Francisco, CA: Jossey-Bass.
La Porta, R., Lopez-de-Silanes, F., Shleifer, A., & Vishny,
R. W. (1997). Trust in Large Organizations. The American Economic
Review, 87(2, Papers and Proceedings of theHundred and Fourth
Annual Meeting of the American Economic Association (May, 1997)),
333-338.
Lewis, J. (2003). Design issues. In J. Ritchie & J. Lewis
(Eds.), Qualitative research practice: A guide for social science
students and researchers (pp. 47-76). Thousand Oaks, CA: Sage.
Li, P. P. (2012). When trust matters the most: The imperatives
for contextualising trust research. Journal of Trust Research,
2(2), 101-106. doi: 10.1080/21515581.2012.708494
Mayer, R. C., Davis, J. H., & Schoorman, F. D. (2006). An
integrative model of organizational trust. In R. M. Kramer (Ed.),
Organizational trust: A reader (pp. 82-108). Oxford: Oxford
University Press.
McEvily, B., & Tortoriello, M. (2011). Measuring trust in
organisational research: Review and recommendations. Journal of
Trust Research, 1(1), 23-63.
Michaels, S., Goucher, N., & McCarthy, D. (2006).
Considering knowledge uptake within a cycle of transforming data,
information and knowledge. Review of Policy Research, 23(1),
267-279.
Nannestad, P. (2008). What have we learned about generalized
trust, if anything? Annual Review of Political Science, 11,
413-436.
PytlikZillig, L. M., Tomkins, A. J., Herian, M. N., Hamm, J. A.,
& Abdel-Monem, T. (2012). Public input methods and confidence
in government. Transforming Government: People, Process and Policy,
6, 92-111.
Rousseau, D. M., Sitkin, S. B., Burt, R. S., & Camerer, C.
(1998). Not so different after all: A cross-discipline view of
trust. Academy of Management Review, 23, 393-404.
Sabatier, P. A., & Jenkins-Smith, H. (1993). Policy change
and learning: An advocacy coalition approach. Boulder, CO: Westview
Press.
Sabatier, P. A., & Weible, C. M. (2007). The advocacy
coalition framework: Innovations and clarifications. In P. A.
Sabatier (Ed.), Theories of the policy process (2nd ed., pp.
189-220). Boulder, CO: Westview Press.
Schoorman, F. D., Mayer, R. C., & Davis, J. H. (2007). An
integrative model of organizational trust: Past, present, and
future. Academy of Management Review, 32(2), 344-354.
Small, M. L. (2011). How to conduct a mixed methods study:
Recent trends in a rapidly growing literature. Annual Review of
Sociology, 37, 57-86.
Soper, D. S. (2013). The free statistics calculators website,
Retrieved Jan 10, 2013, from
http://www.danielsoper.com/statcalc/
Tomkins, A. J., PytlikZillig, L. M., Herian, M. N., Abdel-Monem,
T., & Hamm, J. A. (2010, May 17-20). Public input for municipal
policymaking: engagement methods and their impact on trust and
confidence. Paper presented at the 11th Annual Conference on
Digital Government Research, Public Administration Online:
Challenges and Opportunities, Puebla, Mexico.
Tomkins, A. J., PytlikZillig, L. M., Herian, M. N., & Hoppe,
R. (2012). Trust in government: Findings from public budgeting
consultations. Manuscript submitted for publication.
Trochim, W. M. K. (2005). Research methods: The concise
knowledge base. Cincinnati, Ohio: Atomic Dog.
You, J. (2012). Social trust: Fairness matters more than
homogeneity. Political Psychology, 33, 701-721.
1