Drought Assessment During Dry Season Derived from LANDSAT Imagery Using Amplitude Analysis in Sa Kaeo, THAILAND

Fire regimes in many north Australian savanna regions are today characterised by frequent wildfires occurring in the latter part of the 7-month dry season. A fire management program instigated from 2005 over 24 000 km2 of biodiversity-rich Western Arnhem Land aims to reduce the area and severity of late dry-season fires, and associated greenhouse gas emissions, through targeted early dry-season prescribed burning. This study used fire history mapping derived mostly from Landsat imagery over the period 1990–2009 and statistical modelling to quantify the mitigation of late dry-season wildfire through prescribed burning. From 2005, there has been a reduction in mean annual total proportion burnt (from 38 to 30%), and particularly of late dry-season fires (from 29 to 12.5%). The slope of the relationship between the proportion of early-season prescribed fire and subsequent late dry-season wildfire was ~–1. This means that imposing prescribed early dry-season burning can substantially reduce late dry-season fire area, by direct one-to-one replacement. There is some evidence that the spatially strategic program has achieved even better mitigation than this. The observed reduction in late dry-season fire without concomitant increase in overall area burnt has important ecological and greenhouse gas emissions implications. This efficient mitigation of wildfire contrasts markedly with observations reported from temperate fire-prone forested systems.

Download Full-text

Fire regimes and their correlates in the Darwin region of northern Australia

Pacific Conservation Biology ◽

10.1071/pc070177 ◽

2007 ◽

Vol 13 (3) ◽

pp. 177 ◽

Cited By ~ 5

Author(s):

Owen Price ◽

Bryan Baker

Keyword(s):

Fire History ◽

Ad Hoc ◽

Fire Regime ◽

Negative Impact ◽

Fire Suppression ◽

Landsat Imagery ◽

Fire Regimes ◽

Fire Frequency ◽

Dry Season ◽

Generalized Linear Modelling

A nine year fire history for the Darwin region was created from Landsat imagery, and examined to describe the fire regime across the region. 43% of the region burned each year, and approximately one quarter of the fires occur in the late dry season, which is lower than most other studied areas. Freehold land, which covers 35% of the greater Darwin region, has 20% long-unburnt land. In contrast, most publicly owned and Aboriginal owned land has very high fire frequency (60-70% per year), and only 5% long unburnt. It seems that much of the Freehold land is managed for fire suppression, while the common land is burnt either to protect the Freehold or by pyromaniacs. Generalized Linear Modelling among a random sample of points revealed that fire frequency is higher among large blocks of savannah vegetation, and at greater distances from mangrove vegetation and roads. This suggests that various kinds of fire break can be used to manage fire in the region. The overall fire frequency in the Darwin region is probably too high and is having a negative impact on wildlife. However, the relatively low proportion of late dry season fires means the regime is probably not as bad as in some other regions. The management of fire is ad-hoc and strongly influenced by tenure. There needs to be a clear statement of regional fire targets and a strategy to achieve these. Continuation of the fire mapping is an essential component of achieving the targets.

Download Full-text

Remote sensing of fire regimes in semi-arid Nusa Tenggara Timur, eastern Indonesia: current patterns, future prospects

International Journal of Wildland Fire ◽

10.1071/wf05083 ◽

2006 ◽

Vol 15 (3) ◽

pp. 307 ◽

Cited By ~ 13

Author(s):

Rohan Fisher ◽

Wilfrida E. Bobanuba ◽

Agus Rawambaku ◽

Greg J. E. Hill ◽

Jeremy Russell-Smith

Keyword(s):

Remote Sensing ◽

Prescribed Burning ◽

Landsat Imagery ◽

Great Majority ◽

Fire Regimes ◽

Dry Season ◽

Hotspot Detection ◽

Fire Policy ◽

Eastern Indonesia ◽

Semi Arid

Substantial areas of eastern Indonesia are semi-arid (with a pronounced dry season extending from April to November) with extensive areas of uncultivated vegetation dominated by savanna grasslands and woodlands. These are highly fire-prone, despite high population densities reliant on intensive subsistence agriculture and an official national fire policy that prohibits all burning. To date, no regional studies have been undertaken that reliably assess the seasonal extent and patterning of prescribed burning and wildfire. Focusing on two case studies in east Sumba (7000 km2) and central Flores (3000 km2) in the eastern Indonesian province of Nusa Tenggara Timur, the present paper addresses: (1) the efficacy of applying standard remote sensing and geographic information system tools as developed for monitoring fire patterns in savanna landscapes of adjacent northern Australia, for (2) describing the seasonal patterning of burning at village and broader regional scales in 2003 and 2004. Despite recurring cloudiness, which significantly affected daily fire detection of ‘hotspots’ from Advanced Very High Resolution Radiometer and Moderate Resolution Imaging Spectroradiometer sensors, fire mapping from Landsat imagery was undertaken successfully to reveal: (1) fires burnt an annual average of 29% of eastern Sumba (comprising mostly grassland savanna), and 11% of central Flores (with large forested areas); (2) most fire extent occurred in savanna grassland areas, and significantly also in cultivated lands and small remnant patches of forest; (3) most fire activity occurred under harsh, late dry season conditions; and (4) while the great majority of individual fires were less than 5 ha, some late dry season fires were hundreds of hectares in extent. The potential routine application of different image sensors for fire mapping and hotspot detection is considered in discussion.

Download Full-text

A Machine Learning Approach to Map Tropical Selective Logging

10.1101/451856 ◽

2018 ◽

Author(s):

MG Hethcoat ◽

DP Edwards ◽

JMB Carreiras ◽

RG Bryant ◽

FM França ◽

...

Keyword(s):

Machine Learning ◽

Early Stage ◽

Landsat Imagery ◽

Timber Harvest ◽

Brazilian Amazon ◽

Error Rates ◽

Dry Season ◽

Selective Logging ◽

Omission Error ◽

Low Intensity

AbstractHundreds of millions of hectares of tropical forest have been selectively logged, either legally or illegally. Methods for detecting and monitoring tropical selective logging using satellite data are at an early stage, with current methods only able to detect more intensive timber harvest (>20 m3 ha−1). The spatial resolution of widely available datasets, like Landsat, have previously been considered too coarse to measure the subtle changes in forests associated with less intensive selective logging, yet most present-day logging is at low intensity. We utilized a detailed selective logging dataset from over 11,000 ha of forest in Rondônia, southern Brazilian Amazon, to develop a Random Forest machine-learning algorithm for detecting low-intensity selective logging (< 15 m3 ha−1). We show that Landsat imagery acquired before the cessation of logging activities (i.e. the final cloud-free image of the dry season during logging) was better at detecting selective logging than imagery acquired at the start of the following dry season (i.e. the first cloud-free image of the next dry season). Within our study area the detection rate of logged pixels was approximately 90% (with roughly 20% commission and 8% omission error rates) and approximately 40% of the area inside low-intensity selective logging tracts were labelled as logged. Application of the algorithm to 6152 ha of selectively logged forest at a second site in Pará, northeast Brazilian Amazon, resulted in the detection of 2316 ha (38%) of selective logging (with 20% commission and 7% omission error rates). This suggests that our method can detect low-intensity selective logging across large areas of the Amazon. It is thus an important step forward in developing systems for detecting selective logging pan-tropically with freely available data sets, and has key implications for monitoring logging and implementing carbon-based payments for ecosystem service schemes.

Download Full-text