Assigning dates and identifying areas affected by fires in Portugal based on MODIS data

Abstract

An automated procedure is here presented that allows identifying and dating burned areas in Portugal using values of daily reflectance from near-infrared and middle-infrared bands, as obtained from the MODIS instrument. The algorithm detects persistent changes in monthly composites of the so-called (V,W) BurnSensitive Index and the day of maximum change in daily time series of W is in turn identified as the day of the burning event. The procedure is tested for 2005, the second worst fire season ever recorded in Portugal. Comparison between the obtained burned area map and the reference derived from Landsat imagery resulted in a Proportion Correct of 95.6%. Despite being applied only to the months of August and September, the algorithm is able to identify almost two-thirds of all scars that have occurred during the entire year of 2005. An assessment of the temporal accuracy of the dating procedure was also conducted, showing that 75% of estimated dates presented deviations between -5 and 5 days from dates of hotspots derived from the MODIS instrument. Information about location and date of burning events as provided by the proposed procedure may be viewed as complementary to the currently available official maps based on end-of-season Landsat imagery.

Near- and Middle-Infrared Monitoring of Burned Areas from Space

“…2 months and a much wider area, covering a very large number of scars, and not a single one as in the present study. Regarding the estimated days of burning, 75% of estimated dates in the same study [64] presented deviations less than ±5 days from dates derived from hotspots identified by MODIS.…”

“…The ability of V to discriminate between vegetated and non-vegetated surfaces may be used to build up composites of W free from contamination by clouds, whereas the low values of W associated to burned surfaces suggest generating composites of minimum values of W to discriminate burned areas. Adopting this rationale, and in line with previous work [39,64], discrimination of burned areas was performed using values ( W min ) of a monthly composite of minimum of W and values ( ΔW min ) of…”

“…Results obtained using a similar procedure over the whole territory of Portugal for August and September 2005, one of the worst severe years in terms of burned area, [64] present an overall accuracy of 95.6% and commission and omission errors of 66.5 and 37.1%, respectively. However, the study encompasses a period of Table 3.…”

“…For each burned pixel identified by the algorithm above-described, the date of burning is estimated by analyzing the time series of W for that pixel and searching for the day where maximum temporal separability is achieved [64]. For most cases, time series of W present daily fluctuations of rather small amplitude (Figure 3) which allows identifying the day when the burning event took place by the significant decrease in W following the event.…”

Using cellular automata to simulate wildfire propagation and to assist in fire management

“…Restricted to Portugal, wildfires have burned over 1.1 million ha in the last decade (San-Miguel-Ayanz et al, 2017), and the recent tragic events caused by the megafires of June and October 2017 have left a deep mark at the political, social, economic and environmental levels. Given the increasing trend in both extent and sever-ity of wildfires (Pereira et al, 2005(Pereira et al, , 2013DaCamara et al, 2014;Panisset et al, 2017), the availability of modeling tools of fire episodes is of crucial importance.…”