This study assessed the accuracy of land cover change (2000–2018) maps compiled from Landsat images with either automated digital processing or with visual interpretation for a tropical forest area in Indonesia. The accuracy assessment used a two-stage stratified random sampling involving a confusion matrix for assessing map accuracy and by estimating areas of land cover change classes and associated uncertainty. The reference data were high-resolution images from SPOT 6/7 and high-resolution images finer than 5 m obtained from Open Foris Collect Earth. Results showed that the map derived from automated digital processing had lower accuracy (overall accuracy 73–77%) compared to the map based on visual interpretation (overall accuracy 80–84%). The automated digital processing map error was in differentiating between native forest and plantation areas. While the visual interpretation map had a higher accuracy, it did not consistently differentiate between native forest and shrub areas. Future improvement of the digital map requires more accurate differentiation between forest and plantation to better support national forest monitoring systems for sustainable forest management.

Abstract. Almost every dry season, there are large forest/land fires in several regions in Indonesia, especially in Kalimantan and Sumatra in the dry season of August to September 2015 a forest fire in 6 provinces namely West Kalimantan, Central Kalimantan, South Kalimantan, Riau, Jambi, and South Sumatra. Even some parties proposed that the Government of Indonesia declares them as a national disaster. The low-resolution remote sensing data have been widely used for monitoring the occurrence of forest/land fires (hotspots), and mapping of burnt scars. The hotspot detection was done by utilizing the data of NOAA-AVHRR and MODIS data which have a lower spatial resolution (1 km). In order to increase the level of detail and accuracy of product information, this research is done by using Landsat 8 TIRS (Thermal Infrared Sensor) band which has a greater spatial resolution of 100 m. The purpose of this research is to find and to determine the threshold value of the brightness temperature of the TIRS data to identify the source of fire smoke. The data used is the Landsat 8 of several parts of Borneo during the period of 24 August to 18 September 2015 recorded by the LAPAN's receiving station. Landsat -8 TIRS band was converted into brightness temperature in degrees Celsius, then dots in a region that is considered the source of the smoke if the temperature of each pixel in the region > 43 o C, and given the attributes with the highest temperatures of the pixels in the region. The source of the smoke was obtained through visual interpretation of the objects in the multispectral Natural Color Composite (NCC) and True Color Composite (TCC) images. Analysis of errors (commission error) is obtained by comparing the temperature detected by TIRS band with a visual appearance of the source of the smoke. The result of the experiment showed that there were detected 9 scenes with high temperatures over 43 o C from the 27 scenes Kalimantan Landsat 8 data, which include 153 sites. The accuracy (commission error) of identification results using temperature ≥ 51°C is 0%, temperature ≥ 47°C is 10%, and temperature ≥ 43°C is 30.5%. MODIS data application and its accuracy for detecting point location of a forest fire. Operational of near-real time MODIS temporal series at 1-km spatial resolution to detect forest fire was applied by Ichoku et al., (2004) that mainly using the 4 µm and 11 µm MODIS channels to produce brightness temperature to measure the rate of emission of Fire Radiative Power (FRE) which is the MODIS fire product. Wang et al., (2012) also use a composite of NDVI for a 16-day period to monitor the burnt scars for large scale. Recent method to detect burnt scars using sub-pixel (SPM) of MODIS conducted by Ling et al., (2015) which aim to refine the accuracy of MODIS that in average produce information in 500 m.On the other hand, research on utilizing high spatial resolution to map the burnt scars such as Landsat have also been conducted by several researchers. Jaya (2000) use Landsat TM imagery for identifying the locat...

Landsat-8 has various channels that function to identify an object. The vegetation index algorithm which is based on remote sensing involves several bands and can describe the percentage of canopy and density of vegetation. More than 100 vegetation index algorithms and each can be used in accordance with the research objectives. In this paper we will discuss the utilization of Landsat-8 metric data with the parameters of Normalized Difference Vegetation Index (NDVI) and Normalized Burn Ratio (NBR) and several parameters in metric data with various features to produce indications of rapid land change, especially to detect changes in tree cover area to lose tree cover and vice versa. For this purpose, the annual Landsat-8 metrics data is located in Riau Province. To compare both NDVI and NBR parameters, the trial and error method is used and the results are compared visually to the two different images of the year. The result is that the NBR parameters with a maximum-70 feature and the threshold for tree cover loss and tree cover gain respectively more than -0.1 provide tangible results in looking at the tree cover changes in Riau Province. In the analysis, other information is needed, for example, a map of the Forest Area to see further whether the changes that occur are in the forest area or not, which will certainly provide different treatment.