Publication Detail

Immediate Resource Requirements after Hurricane Katrina

UCD-ITS-RP-12-98

Reprint

Suggested Citation:
Holguín-Veras, José and Miguel Jaller (2012) Immediate Resource Requirements after Hurricane Katrina. ASCE’s Journal Natural Hazards Review 12 (2), 117 - 131

The paper focuses on the quantitative study of immediate resource requirements, i.e., the needs that arise in the aftermath of a disaster, which is one of the most severely understudied aspects of humanitarian logistics. The paper develops numerical estimates of these requirements and their temporal patterns using a data set put together by postprocessing the requests made by emergency responders in the aftermath of Hurricane Katrina. The analyses of the data provide a glimpse into what was needed at the site and when. A key finding is that a relatively small number of different items, approximately 150, were requested, which is a fraction of previous estimates that suggested 350–500 different commodities. The analyses reveal that an even smaller number of items concentrated the bulk of the requests: 20 commodities accounted for approximately 30% of the requests, 40 commodities for 47%, and 50 commodities for 56%. The relatively low number of different items makes it easier to preposition these key supplies or to develop regional blanket purchase agreements to promptly obtain them from preapproved vendors. These measures are cost-effective ways to reduce delivery times, as a relatively small initial investment could cover a large portion of the needs. The paper explored the use of statistical models to depict the temporal patterns of requests. As part of the research, autoregressive integrative moving average (ARIMA) models were estimated for the key commodity groups that provide a solid analytical framework for forecast. The fact that it is possible to estimate robust ARIMA models to forecast resource requirements has important implications because it opens the door for the combined use of need forecast, inventory control, and ordering models. Given the lead times between request and arrival of supplies, the integration of forecasts into the ordering process will translate into a better match between supply and demand.