Policy discussion and conclusions - A Difference-in-Differences Analysis in Finland

A Difference-in-Differences Analysis in Finland

5. Policy discussion and conclusions

A major theme that appears through the tested scenarios is that constraining urban development in flood risk areas produces urban morphologies that are more fragmented relative to the baseline—

indicated by an increase of the amount of urban clusters and concurrent reduction of their size—

irrespective of whether the interventions stem from a planning system that adjusts to market forces (MR scenarios) or constrains those forces (DR scenarios). This implies that green spaces are more fragmented as well, with a larger proportion of total built-up land exposed or proximate to ecosystem services. So, it can be maintained that planning interventions that restrict growth patterns in flood risk areas will slow down the consolidation of urban areas characterizing the BAU scenario, which combined with reduced growth rates may encourage the preservation and spatially heterogeneous presence of flood-regulating and other ecosystem services. As discussed in Section 4.1, alleviating the loss of ground-based ecosystem services that are embedded in the urban tissue reduces the loss of significant amounts of wealth in the real estate market (thus reducing vulnerability), while increasing the exposure of coastal residential areas to ecosystem services embedded in the urban tissue.

While the scenario simulations indicate a halt in varying indicators of growth, only the main driver of growth as modelled by SLEUTH (in this case, edge growth – see Section 2.4) is affected. The other types of urban growth simulated by SLEUTH are not affected by the scenarios (diffusive, road-influenced, and new spreading center growth). This implies that constraining growth in risk areas

areas, but does not change the underlying urban growth drivers as modelled by SLEUTH. The halting of growth rates inside risky areas in combination with the morphological benefits can be at first assumed, as in the previous paragraph, as beneficial for the coastal zone, but as discussed in section 4.2 this not the entire story and indirect effects must also be accounted for. Although the simulated changes in growth rates are rather weak, it is important to understand how the impacts of these deviations from the baseline are distributed across urban economic sectors.

An interesting comparison between policy instruments that constrain market forces versus ones that respond to those forces has also been argued in Sections 4.2 and 4.3. The simulations indicate that different land constraints can have a differential redistribution of development activity in and near the area of application, and that demand for amenity-rich but safe locations will not always translate to actual refocusing of development. In this respect, the spatial character of interventions becomes important, as interventions that track and respond to market adjustments caused by increasingly transparent climate-related risks appear to be a necessary element in refocusing urban development.

The tolerance of the planning system to the levels of flood risk and to market behavior is therefore a parameter in the way wealth and investment in the form of capital stock and infrastructure are distributed in relation to climate-sensitive risks and amenities so as to better reflect the spatial configuration of risks. Based on the results, however, it is unclear whether a planning intervention fully following market responses is preferable over one that poses ad hoc but gentle restrictions according to flood risks. In any case, the simulations can be taken as an indication that a combination of market-led/information-based and zoning-based regulatory elements can provide the necessary precision and agility for a flood-related adaptation strategy.

It is also noteworthy that a policy that excludes the entire floodplain from future development translates to reductions between 25 and 40% in the produced built-up land relative to the baseline.

On one hand, as discussed in Section 4.1, these elasticities are indicative of the volume of development that would occur in the floodplain without any intervention. On the other hand, when looking at the scenario performances inside the various flood risk zones, it is chiefly the fully restrictive scenario that confirms the fact that a full exclusion of the flood plain from future development irrespective of market forces is capable of subduing a tremendous amount of growth, whereas all other scenarios that take less restrictive approaches appear to induce results quite near to each other, irrespective of the method used to quantify the development restrictions. This may strengthen the view expressed previously, namely development restrictions that are spatially flexible, rather than monolithic throughout the floodplain, may be able to re-distribute growth more elegantly without inducing a shock with magnitudes that intuitively appear too problematic for the urban development. It also begs the question of where the subdued growth is rechanneled to, which brings us to a rather complex issue. Additional tests need to be performed in order to understand how SLEUTH handles development potential that is realized in a baseline scenario with certain land constraints, but is unrealized in scenarios that impose additional (to the baseline) land constraints. It would be particularly interesting to see whether alternative land use change models would spatially redistribute growth in an altogether different altogether manner than wat is presented in this analysis.

The problem is in fact a difficult one, and it further relates to the ability of models informed by complexity theory to calculate the development potential of an urban area before and independently

edges relative to the baseline and re-bounces to positive after about 2030 with about 0.1% more edges.

Concerning local effects (Figure 9), a noteworthy feature of the DR simulations is that a flat-out exclusion policy for the entire floodplain, as represented by scenario DRb, yields a–1.6% deviation from the baseline in the production of built-up land in the coastal flood-safe zone (300 m from the coast), while the rest of the scenarios have similar deviations to scenario MRa of about –0.5% relative to the baseline. The underlying exclusion-attraction values in this flood-safe area are the same for all scenarios (neutral value of 50) except MRb. Compared to scenario MRb, which actively redistributed development in the same area, this ties in with what was indicated previously that—given a modelling approach that assumes that market forces are inherent in the end result of simulations—demand for flood-safe locations will not automatically translate to refocusing of development. The additional suggestion here is that a planning system that is entirely insensitive to different flooding probabilities will induce disproportional changes in areas that are communicated as safe from risks.

Lastly, all scenarios have near-zero deviation from BAU in the flood-safe areas between 0.3 and 1 km from the coast, and the differences reappear in the flood-safe areas between 1 and 10 km; these two zones have identical land constraints in all scenarios, whereas in inland flood-safe areas between 1 and 10 km from the coast, the effects of the various reappear. Although interesting, this feature cannot be explored with the current model setup. While potentially connected to the spatial interaction effect of growth constraints in the exclusion-attraction layer that was discussed earlier, a closer look than afforded by this study is needed mainly on the way SLEUTH models how growth potential in the entire modeled area is affected by imposed restrictions in particular areas.

5. Policy discussion and conclusions

A major theme that appears through the tested scenarios is that constraining urban development in flood risk areas produces urban morphologies that are more fragmented relative to the baseline—

indicated by an increase of the amount of urban clusters and concurrent reduction of their size—

While the scenario simulations indicate a halt in varying indicators of growth, only the main driver of growth as modelled by SLEUTH (in this case, edge growth –see Section 2.4) is affected. The other types of urban growth simulated by SLEUTH are not affected by the scenarios (diffusive, road-influenced, and new spreading center growth). This implies that constraining growth in risk areas

areas, but does not change the underlying urban growth drivers as modelled by SLEUTH. The halting of growth rates inside risky areas in combination with the morphological benefits can be at first assumed, as in the previous paragraph, as beneficial for the coastal zone, but as discussed in section 4.2 this is not the entire story and indirect effects must also be accounted for. Although the simulated changes in growth rates are rather weak, it is important to understand how the impacts of these deviations from the baseline are distributed across urban economic sectors.

It is also noteworthy that a policy that excludes the entire floodplain from future development translates to reductions between 25 and 40% in the produced built-up land relative to the baseline.

On one hand, as discussed in Section 4.1, these elasticities are indicative of the volume of development that would occur in the floodplain without any intervention. On the other hand, when looking at the scenario performances inside the various flood risk zones, it is chiefly the fully restrictive scenario that confirms the fact that a full exclusion of the flood plain from future development irrespective of market forces is capable of subduing a tremendous amount of growth, whereas all other scenarios that take less restrictive approaches appear to induce results quite near to each other, irrespective of the method used to quantify the development restrictions. This may strengthen the view expressed previously, that development restrictions that are spatially flexible, rather than monolithic throughout the floodplain, may be able to re-distribute growth more elegantly without inducing a shock with magnitudes that intuitively appear too problematic for urban development. It also begs the question of where the subdued growth is rechanneled to, which brings us to a rather complex issue. Additional tests need to be performed in order to understand how SLEUTH handles development potential that is realized in a baseline scenario with certain land constraints, but is unrealized in scenarios that impose additional (to the baseline) land constraints.

It would be particularly interesting to see whether alternative land use change models would spatially redistribute growth in an altogether different manner than what is presented in this analysis.

The problem is in fact a difficult one, and it further relates to the ability of models informed by complexity theory to calculate the development potential of an urban area before and independently

addressed by cellular automata models. The standard approach in urban and regional research is to look into microeconomic theory, which explains how total regional and national economic output is distributed over an urban area through new investments and growth in capital stock and infrastructure, based on the location decisions of firms and households. This is beyond the scope of this analysis, but future research would certainly need to better relate cellular automata models to the way urban microeconomic theory explains why cities exist and how they evolve as they do.

Regardless of the uncertainties that are common in any empirical approach, translating econometric estimations to scenario input has been a straightforward task, and the defining parameter in this task is the particular way the statistical estimations are translated to pixel values. This paper offered one possible translation, but there are undeniably other approaches. The modelling and forecasting capacity with respect to spatially disaggregate dynamics is SLEUTH’s main strength, because this kind of information is typically left unaccounted for in most adaptation studies, resulting in partial adaptation knowledge that is missing essential spatial parameters. Such output is of ever-increasing importance; as the need for more context-aware strategies rises, SLEUTH can contribute on one hand to vulnerability and exposure assessments, and on the other hand to physics/engineering-based hazard research.

In closing, it is important to note that SLEUTH’s distinguishing feature in helping to navigate through alternative urban futures is its ability to simulate the distribution of urban growth at a fine geographical grid. This benefit will be boosted if coupled to a model that assesses monetary costs and benefits, which at first means understanding the impact of SLEUTH’s forecasts to indicators external to the model. For instance, if growth is constrained or promoted in a part of the city, and SLEUTH translates this strategy into a likely urban morphology, it would be vital to know how (i) the forecasted growth will impact house prices, mobility patterns, redevelopment of the existing building stock, ecosystems’ composition, urban microclimate and so on, and (ii) how this strategy can be efficiently achieved in the first place through investment, taxation, or other interventions, and what would these interventions mean for the welfare of households and firms. As such tasks are beyond the scope of SLEUTH, it should be stated that SLEUTH, with its rare ability to model urban growth in a spatially disaggregate manner, can serve as a link between various other models that focus on the behavior of multiple urban economic sectors (urban microeconomic models; integrated land use transport models;

regional CGE models) but cannot distribute growth in a fine resolution grid as SLEUTH does.

6. References

Aerts J.C.J.H., Wouter Botzen W.J., Emanuel K., Lin N., de Moel H., Michel-Kerjan E.O. (2014), Evaluating Flood Resilience Strategies for Coastal Megacities, Science 344: 473–475.

Aguejdad R., Hidalgo J., Doukari O., Masson V., and Houet T. (2012), Assessing the influence of long-term urban growth scenarios on urban climate, 6th International Congress on Environmental Modelling and Software, Leipzig, Germany.

36(3): 1426–1464.

Barneveld H.J., Silander J.T., Sane M., and Malnes E. (2008), Application of satellite data for improved flood forecasting and mapping. In 4^th International Symposium on Flood Defence:

Managing Flood Risk, Reliability and Vulnerability, Toronto, Ontario, Canada (pp. 77-1–77-8).

Batty M. (1997), Cellular Automata and Urban Form: A primer, Journal of the American Planning Association 63 (2) pp. 266–274.

Batty M. (2007), Cities and Complexity: Understanding Cities with Cellular Automata, Agent-Based Models, and Fractals, The MIT Press, Cambridge MA.

Bin O., Crawford T.W., Kruse J. B. and Landry C.E. (2008a), Viewscapes and flood hazard: coastal housing market response to amenities and risk, Land Economics 84(3): 434–448.

Bin O., Krusc J.B., and Landry, C.E. (2008b), Flood hazards, insurance rates, and amenities: evidence from the coastal housing market, The Journal of Risk and Insurance 75(1): 63–82.

Boesch P., Ciari F., and Perrels. A (2014), D2.3 – Overview of system responsiveness to climate change impacts in energy, transport and tourist sectors, EU FP7 TOPDAD project, http://www.topdad.eu/upl/files/102695.

Brander L.M. and Koetse M.J. (2011), The value of urban open space: Meta-analyses of contingent valuation and hedonic pricing results, Journal of Environmental Management 92: 2763–2773.

Caglioni M., Pelizzoni M., and Rabino G.A. (2006), Urban Sprawl: A Case Study for Project Gigalopolis Using SLEUTH Model, Lecture Notes in Computer Science 4173: 436–445.

Candau J.T. (2002), Temporal Calibration Sensitivity of the SLEUTH Urban Growth Model, Master’s thesis, University of California Santa Barbara.

Chaudhuri G. and Clarke K.C. (2013), The SLEUTH Land Use Change Model: A Review, The International Journal of Environmental Resources Research 1(1): 88-104.

Clarke K.C. (2008), A Decade of Cellular Urban Modeling with SLEUTH: Unresolved Issues and Problems, in: R.K. Brail (Ed.) Planning Support Systems for Cities and Regions, Lincoln Institute of Land Policy, Cambridge MA, 47–60.

Clarke K.C., Gaydos L., and Hoppen S. (1997), A self-modifying cellular automaton model of historical urbanization in the San Francisco Bay area, Environment and Planning B 24: 247–261.

Clarke K.C. and Gaydos L. (1998), Loose-coupling a cellular automaton model and GIS: Long-term urban growth prediction for San Francisco and Washington/Baltimore, International Journal of Geographic Information Science 12: 699–714.

Daniel V.E., Florax R.J.G.M., and Rietveld P. (2009), Flooding risk and housing values: an economic assessment of environmental hazard, Ecological Economics 69(2): 355–365.

In document Space and Price in Adapting Cities : Exploring the Spatial Economic Role of Climate-Sensitive Ecological Risks and Amenities in Finnish Housing Markets (sivua 134-162)