Forest Conversion in Upstream Oil and Gas Industry Area

. Forest conversion is the root of the problem that can hamper investment, state revenue, and potentially cause state losses. The purpose of this study was to analyze forest conversion in the upstream oil and gas industry area and the factors that influence forest conversion in the upstream oil and gas industry area based on spatial analysis. This study used systematic sampling with the point grid method measuring 1 kilometers * 1 kilometers. Rokan Block had 5,125 sampling points. This study used raster-based GIS analysis and logistic regression with the Wald Test. The results of the study showed that the Rokan Block Area experienced a fairly severe forest conversion in 2000-2014 influenced by the total population and The 1986 TGHK. The novelty of this study is that the active role of communities around the forests that still upholds the concern for the forest must be preserved, because they will be the front guard in securing forest encroachment, and upstream oil and gas industry are not the main economic activities that cause forest conversion.


INTRODUCTION
The oil and gas industry is an important sector and contributed 66 percent of total non-tax revenue (PNBP) and absorbed a large workforce. [ 1 ]. Rokan Block Riau Province contributed 40 percent of national oil demand and 30 percent of regional income, [ 2 ], but there is an opinion that the upstream oil industry is one of the factors causing land-use change from forests to non-forests (forest conversion). Harun, [ 3 ] stated that the upstream oil industry is one of the main causes of forest conversion in Riau. Harun  Land-use change in the implementation of development cannot be avoided. These changes occur because of two things, namely the need to meet the increasing needs of the population and increasing demands for a better quality of life. McNeil, [ 5 ] stated that the driving factors of land-use change are politics, economics, demography, and culture. Land-use change is a reflection of human efforts in utilizing and managing land resources that have an influence on humans and environmental conditions.
The main foundation for the implementation of spatial use is the spatial planning document as legislation that binds the public and also government officials. However, spatial plans in Indonesia in the field are not implemented with the same level of discipline as zoning documents in the regulatory system. [ 6 ]. The purpose of this study was to analyze forest conversion in the upstream oil industry area and the factors that influence forest conversion in the upstream oil industry area with spatial analysis.
Spatial analysis is a quantitative method to see the diversity of things spatially. Geographic information system (GIS) is an automation system for handling spatial data. This system can encapsulate information intelligence geographically (spatial). In spatial modeling, there are two categories of area data structures namely vector and raster. Vector is a data structure based on coordinates and shapes such as points, lines or areas, while raster is a data structure based on cells, such as remote sensing satellite imagery data.

II. METHODS
The population in this study was the total land area in the Rokan Block. Rokan Block has an area of 650,000 Ha. Data on these variables were processed in the form of grid/raster data. With a resolution of 30 meters * 30 meters per pixel. Sample selection was performed by using systematic sampling based on the method of a grid point by Jensen. [ 8 ]. The point used for the sample was 1 kilometers * 1 kilometers. The illustration of systematic sampling can be seen in Figure 1.  Distance from oil and gas well X6 Total population X 12 The 1986 TGHK β0 = Constants β1-n = 1 st to n-th independent variable coefficient X1-n,i = 1 st to n-th independent variables on i dependent variable n = Number of variables Wald test is used to examine the effect of independent variables on the dependent variable partially. Wald test works by comparing the Wald statistical value with the Chi square comparative value at degrees of freedom (db) = 1. In this test the authors used a significance value (ϼ) to determine whether or not the independent variable affects the dependent variable. If ϼ-value is smaller than α, it can be concluded that the independent variable can partially influence the dependent variable. Reserve and Tourism Forest, Limited Production Forest, Permanent Production Forest, and Convertible Production Forest. In addition to the 5 classifications, there are 2 nonforestry function classifications namely Non-Forest Area (APL) and body of water. The third point of SK.173/kpts-II/1986 instructs the Head of the Forest Inventory and Utilization Agency to carry out the measurement and arrangement of the forest area boundaries in the field but is not implemented in most forest areas in Riau Province. This has become one of the factors in violation of the 1986 TGHK. In addition, weak governance and oversight also contributed to the ineffectiveness of the 1986 TGHK in spatial planning so that many violations were found. Many of the areas designated as forest areas in The 1986 TGHK have been converted to land uses outside the forestry function. The 1986 TGHK established 97.8 percent of Riau Province as a forest area. For Rokan Block, 91% of the Rokan Block area is a forest area such as Protected Forest, Nature Reserve and Tourism Forest, Production Forest, Convertible Production Forest, and Limited Production Forest as shown in Table 1   In addition, several processes such as: a. Detection of forest loss due to fire; b. Detection of smallholder rotational agriculture in tropical forests; c. Detection of selective logging; d. Detection of short cycle plantations in subtropical and tropical zones.
Hansen et al. [ 9 ] classified forest data and its changes in a multi-temporal manner. The distribution of forest in Rokan Block in 2000 that will be used for the study needs to be converted into dummy data (coding of variables to 0 and 1), as in Figure 1 below.

Figure 1. Land Cover Map of Rokan BlockForest Area in 2000 (Dummy)
The forest area in Rokan Block in 2000 as a dummy result was 521 thousand hectares (5.8 million pixels) or around 80 percent of Rokan Block. The 80 percent forest area is a sizeable proportion compared to the non-forest area which is only about 20 percent as shown in Table 2. When compared to the 1986 TGHK plan, the forest area in Rokan Block was 91 percent, so that in 2000 there was a difference forest area of around 11 percent.

Forest Conversion in 2000-2014
The data source used to see the forest conversion in 2000-2014 was dummy analysis of Global Forest Change data from 2000-2014. Forest loss was defined as a disturbance in the replacement of stands or changes from forest to non-forest

LAND COVER MAP OF ROKAN BLOCK FOREST AREA
conditions. The description of forest conversion in 2000-2014 occurred over a large area as can be seen in Figure 2 below.  Table 3. Margono [ 9 ] stated that in this period, Riau experienced the largest forest conversion in the world. The period above 2000 is the period after the regional autonomy policy recommended through Resolution of MPR RI Number IV/MPR/2000 on Policy Recommendations in the Implementation of Regional Autonomy. Many people assume that the implementation of regional autonomy policy also influences forest conversion.

Influencing factors of forest conversion
The results of Wald Test showed that almost all independent variables X had a significant influence on the dependent variable Y with Sig.<0.05, except for two variables namely X6 (Total Population) and X13 (The 1986 TGHK) with sig ≥ 0.05, which showed that both variables had no significant influence on forest conversion as in Table 4 below. The proportion of contributions to Rokan Block was performed on variables that have been known to influence the previous analysis. The proportion of contributions was performed with 3 testing groups. The first group was conducted on biophysical component variables namely: distance from river (X1), rainfall (X2), type of soil (X3), elevation (X4) and slope (X5). The second group was performed on social variables: distance from settlement (X7), distance from road (X8) and distance from government office(X9).The third group was performed on economic variables : distance from HTI (X10), distance from HGU (X11), and distance from oil and gas well(X12).
The teston the biophysical variable group showed that the type of soil (X3) was the most influential variable and contributed significantly to the increase in R-Square with a high proportion of increase. Meanwhile, the contribution of other variables respectively was slope (X5), elevation (X4), rainfall (X2), and distance from the river (X1), as shown in Figure 3 below. The test on the social variable group showed that distance from settlement (X7) was the most influential variable and contributed significantly to the increase in R-Square with a high proportion of increase. Meanwhile, the contribution of other variables respectively was distance from government office (X9) anddistance from road (X8) as shown in Figure 4 below. The test on the economy variable group showed the distance from HGU (X11) which contributed significantly to the increase in R-Square with a proportion of an increase of almost 15%. Meanwhile, distance from oil and gas well (X12) made the smallest contribution between HGU (X11) and HTI (X10), as shown in Figure 5.

IV. CONCLUSION
Rokan Block Area experienced a pretty bad forest conversion in 2000-2014. During this period, forest cover in Rokan Block declined dramatically from 80 percent in 2000 to 42 percent in 2014.
Forest conversionwas significantly influenced by The insignificant of the 1986 TGHK showed weak governance and implementation of spatial planning policies that caused forest conversion to be out of control in 2000-2014.
Forest conversion practitioners prefer forests on mineral soils as a priority because of easy access. The active role of communities around the forests that still uphold the concern for the forest must be preserved, because they will be the front guard in securing forest encroachment, and oil and gas are not the main economic activities that cause forest conversion.

V. ACKNOWLEDGMENTS
We would like to express my gratitude to Universitas Riau and PT.Chevron Pacific Indonesia for the assistance so that this research could be carried out.