GUEST POST: New maps reveal more complex picture of Sumatran fires
By David Gaveau
Extreme episodes of trans-boundary haze in Southeast Asia in 2013 and 2014 — and an anticipated El-Niño-induced drought during the second half of 2014, which could result in significantly heightened fire activity across Indonesia — have focused attention on causes and origins of fires in peatland areas of Sumatra. It has also prompted the affected countries in the region to develop policies to mitigate future fires in the region — as well as penalties for those who start the fires. Indonesia’s vice president has convened senior ministers from several agencies to combat the problem, and a special “situation room” is being established to ensure firefighting capabilities and response within hours after new fire hotspots are detected by satellites. In parallel, Singapore has drafted a bill that would allow it to fine companies for fires that take place on their plantations. (Haze wafting from Indonesian fires in June 2013 choked the skies of Singapore, causing record-high levels of air pollution.) Singaporean authorities, as well as many advocacy groups, now blame plantation companies (large-scale operators) for the recent burning, based upon analyses of fire hotspots and cadastral maps showing that over half of hotspots during the recent 2013 and 2014 fire event in Sumatra was within concessions — i.e., land allocated to companies for plantation development of industrial-scale monoculture oil palm and Acacia. Policy interventions for mitigating fires require solid information on who is setting fires to whose land. However, this basic information is missing, because competing claims over land ownership in Indonesia gives rise to confusion over who is setting fires to whose land. Land use and land tenure in Indonesia are governed by a tangle of national, provincial and customary laws that often compete with each other, resulting in confusion over who owns which bit of land. This situation is exacerbated by an influx of land-seeking migrants and by investments in agricultural expansion by mid-level investors of unknown origin. Tensions — and occasionally conflict — can arise among these land users. Another confounding factor is the fact that fires move across the landscape, propelled by topography and wind. One cannot assume — based only on fire hotspot locations overlaid on concession maps — that burning within concessions is caused by large companies. There are two main reasons for this:
- Land within a concession boundary may not be fully controlled by the corporation in question. A large part may be claimed and occupied by unrelated land users, and these land users may be the ones who start fires in the concessions.
- Fires can start outside the concessions and spread into the concessions.
We show here that it is possible to produce detailed maps of land use in concessions, showing the areas occupied by unrelated land users (small-scale and mid-level operators) using before-and-after fire images from NASA’s recently launched LANDSAT satellite. We can also produce detailed maps of fire progression, showing where fire starts and where fire ends, using a combination of LANDSAT and fire hotspots data. We applied our methods to the February and March 2014 fires that generated the two largest burned areas in Riau province, Sumatra — one in Bengkalis district and one on the nearby island of Pulau Rupat. Our analysis reveals that the fires either started outside concessions, or on land occupied by small-scale operators (local or migrant communities) within the concessions (see Figure 1 at top, and Figure 2, below).
While our analysis should not be read as completely absolving companies — as companies have in the past used fire to clear land — it does show that the situation on the ground is more complex than is often portrayed, and that companies are not always entirely to blame when fires are started on their concessions. Our findings are likely to complicate efforts to hold companies accountable and to penalize offenders. Our analysis points to the need to continue — and expand — this type of monitoring, and to move away from assigning blame simply by overlaying hotspot and concession maps. The situation on the ground, as we now can see, is more complex than that.