Increasing prices and demand for biofuel and cooking oil from importer countries have caused a remarkable expansion of oil palm plantations in Indonesia. In this paper, we attempt to monitor the expansion of oil palm plantations on peat land and in tropical forests. We measure the GHG emissions from the land conversion activities at provincial scale. Using Landsat images from three different periods (1990s, 2000s and 2012), we classified LULC of the Riau Province, which is the largest oil palm producing region in Indonesia. A hybrid method of integration, generated by combining automatic processing and manual analysis, yields the best results. We found that the tropical rainforest cover decreased from ∼63% in the 1990s to ∼37% in the 2000s. By 2012, the remaining tropical rainforest cover was only ∼22%. From the 1990s to the 2000s, conversion of forests and peat lands was the primary source of emissions, total CO₂ emitted to the atmosphere was estimated at ∼26.6 million tCO₂.y^-1, with 40.62% and 59.38% of the emissions from conversion of peat lands and forests, respectively. Between 2000 and 2012, the total CO₂ emitted to the atmosphere was estimated at ∼5.2 million tCO₂. y^-1, with 69.94% and 27.62% of the emissions from converted peat lands and converted forests, respectively. The results show that in the Riau Province, the oil palm industry boomed in the period from 1990 to 2000, with transformation of tropical forest and peat land as the primary source of emissions. The decrease of CO₂ emissions in the period from 2000 to 2012 is possibly due to the enforcement of a moratorium on deforestation.

The previous studies of upwelling within the Indonesian seas only focused on the Southeast Monsoon (SEM) season. The western coast of Sumatra Island; the southern coast of Java Island, Sunda Islands, and Sulawesi Island; the Banda Sea, the Maluku Sea, and the Arafura Sea are well-known SEM upwelling areas. However, the upwelling events during Northwest monsoon (NWM) have never been investigated. The investigation of NWM upwelling is challenging due to the broad cloud coverage, limiting the observation of infrared and visible sensors of the satellites. This study used the blended products of the satellite-based sea surface temperature (SST), chlorophyll-a (chl-a), sea level anomaly (SLA), surface wind, and rainfall to identify the NWM upwelling occurrences. Remote sensing reflectances 443 nm and 555 nm were also used to examine the influence of suspended sediment and organic matter which may bias the chl-a concentration. Along the northern coast of the Lesser Sunda Islands, the north coast of Papua Island and the Malacca Strait are the areas of NWM upwelling as denoted by the positive anomaly of chl-a, the negative anomaly of SST, and the negative anomaly of SLA. Focusing on the area along the northern coast of the Lesser Sunda Islands, the NWM upwelling is generated by offshore Ekman Mass Transport that reduces SST and increases chl-a. From 2010 onwards, El Niňo Southern Oscillation has a more consistent effect on the NWM upwelling than Indian Ocean Dipole. El Niňo (La Niňa) tends to weaken (strengthen) upwelling. The magnitude of NWM upwelling is weaker than SEM upwelling.

Background: The availability of urban forest is under serious threat, especially in developing countries where urbanization is taking place rapidly. Meanwhile, there are many classifier algorithms available to monitor the extent of the urban forest. However, we need to assess the performance of each classifier to understand its complexity and accuracy. Methods: This study proposes a novel procedure using R language with RStudio software to assess four different classifiers based on different numbers of training datasets to classify the urban forest within the campus environment. The normalized difference vegetation indices (NDVI) were then employed to compare the accuracy of each classifier. Results: This study found that the Extreme Gradient Boosting (XGBoost) classifier outperformed the other three classifiers, with an RMSE value of 1.56. While the Artificial Neural Network (ANN), Random Forest (RF), and Support Vector Machine (SVM) were in second, third, and fourth place with RMSE values of 4.33, 6.81, and 7.45 respectively. Conclusions: The XGBoost algorithm is the most suitable for urban forest classification with limited data training. This study is easy to reproduce since the code is available and open to the public.

Identification and monitoring of species composition and richness is needed to formulate effective mangrove management and conservation priorities. Prior studies have used commercial satellite images which are cost prohibitive for national and global applications. Here, we used freely available Landsat satellite data and new indices to discriminate mangrove species in Maros Regency, South Sulawesi, Indonesia and Segara Anakan, West Java, Indonesia. We use sensitive algorithm of the principal polar spectral (PPS) indices to discriminate mangroves species. PPS Indices were produced from a set of 3-dimensional Landsat 8 Operational Land Imager (OLI) spectral indices (PPS Brightness, PPS Greenness, and PPS Wetness) determined by a polar change of the principal component axes of a spectral image of reference scene. We qualitatively compare this set of PPS indices with the set of conventional RGB multi-bands image composition and conventional Normalized Difference Vegetation Indices (NDVI) for mangroves species discrimination. The comparisons indicate that the set of PPS indices have the potential for regional and possibly global applications in mangroves species mapping and monitoring.