Identifying Spatial Patterns of Recovery and Abandonment in the Post-Katrina Holy Cross Neighborhood of New Orleans

Article excerpt

Introduction

In August and September 2005, Hurricanes Katrina and Rita (from this point on referred to as Katrina) made landfall, leaving in their wake massive loss of life, property, business, and community (Cefalu et al. 2006; Hartman and Squires 2006; Kunreuther 2006; Madrid et al. 2006). Unfortunately, as of the summer of 2009, many neighborhoods in Orleans and St Bernard Parish were still struggling to reestablish themselves, the few returnees being faced with multiple abandoned and overgrown properties, lost neighbors and churches (and therefore social support networks), and severely limited infrastructure (Holzer and Lerman 2006; Penner and Ferdinand 2009). This disaster has illustrated existing disparities between communities in New Orleans, and how these social impediments also become hindrances to recovery (Vale and Campanella 2005; Bullard and Wright 2009). We should also acknowledge how little we understand about the recovery process from a spatial perspective. There have been few data-driven spatial analyses at the street and neighborhood scale of who returns, why they return, and what patterns result. This paper will not provide explanations or causations of the fine-scale processes involved but rather suggest a methodological framework needed to begin the discourse.

The Spatial Analysis of Recovery

Although post-disaster recovery has received attention by academics (Haas et al. 1977; Wright et al. 1979; Comerio 1998; Alesh et al. 2009), fine-scale spatial analysis of the processes involved has been lagging for a number of reasons, though mainly because data are not widely available (Duval-Diop and Rose 2008; Mills 2008). Utility or mail service delivery data and other proprietary data collected by contractors could provide spatial insight in some cases, but they are extremely hard to obtain, resulting in difficulties in performing confirmatory analyses by researchers. Consequently, advances or insights cannot be built upon, or even translated to other locations. Other data, such as building permits hold more promise as an accessible point-level source for change. However, all of these data only tell one part of the return/abandonment story; they do not describe the visual impact on neighborhoods and the stress returnees have to deal with.

[FIGURE 1 OMITTED]

A second problem is the spatial scale at which data are available. There are some excellent data reporting bodies and warehouses associated with Katrina and New Orleans, such as the Greater New Orleans Community Data Center and the LSU Clearinghouse Cooperative (Mills et al. 2008). However they do not release finescale residential data for confidentiality reasons. Data released by zip code, by census enumeration unit (Liu and Plyer 2008), or even for an entire "neighborhood" such as the Lower 9th Ward have severe limitations (Duval-Diop et al. 2010; Grubesic et al. 2006), not least of which is that nuances in disaster exposure (e.g., depth of flooding) could be missed at the neighborhood scale. For example, zip code 70177 shown in Figure 1 contains parts of eight different, commonly referenced "neighborhoods," including the Lower 9th Ward and Holy Cross, the subject of this paper (Curtis 2008). Not having fine-scale data is problematic as many of the local processes either impeding or promoting recovery work at the sub-neighborhood level (for example a church stimulating the recovery in the block surrounding it). It is therefore imperative to be able to analyze abandonment and recovery at a finer scale, ideally by home and street segment.

Also, in order to be an effective measure of neighborhood change, data must be collected during multiple time periods as recovery is longitudinal and not just cross-sectional. In order to develop a fine-scale spatial recovery theory, data need to be available at an appropriate scale, for multiple time periods, and with easy accessibility for both researchers and community groups alike. …