*Article* **Assessing the Distribution of Heavy Industrial Heat Sources in India between 2012 and 2018**

**Caihong Ma 1,2,3, Zheng Niu 1,2, Yan Ma 2,\*, Fu Chen 2, Jin Yang <sup>2</sup> and Jianbo Liu <sup>2</sup>**


Received: 28 October 2019; Accepted: 8 December 2019; Published: 10 December 2019

**Abstract:** The heavy industry in India has witnessed rapid development in the past decades. This has increased the pressures and load on the Indian environment, and has also had a great impact on the world economy. In this study, the Preparatory Project Visible Infrared Imaging Radiometer (NPP VIIRS) 375-m active fire product (VNP14IMG) and night-time light (NTL) data were used to study the spatiotemporal patterns of heavy industrial development in India. We employed an improved adaptive K-means algorithm to realize the spatial segmentation of long-term VNP14IMG data and artificial heat-source objects. Next, the initial heavy industry heat sources were distinguished from normal heat sources using a threshold recognition model. Finally, the maximum night-time light data were used to delineate the final heavy industry heat sources. The results suggest, that this modified method is a much more accurate and effective way of monitoring heavy industrial heat sources, and the accuracy of this detection model was higher than 92.7%. The number of main findings were concluded from the study: (1) the heavy industry heat sources are mainly concentrated in the north-east Assam state, east-central Jharkhand state, north Chhattisgarh and Odisha states, and the coastal areas of Gujarat and Maharashtra. Many heavy industrial heat sources were also found around a line from Kolkata on the Eastern Indian Ocean to Mumbai on the Western Indian Ocean. (2) The number of working heavy industry heat sources (NWH) and, particularly, the total number of fire hotspots for each working heavy industry heat source area (NFHWH) are continuing to increase in India. These trends mirror those for the Gross Domestic Product (GDP) and total population of India between 2012 and 2017. (3) The largest values of NWH and NFHWH were in Jharkhand, Chhattisgarh, and Odisha whereas the smallest negative values, the *Slope*\_*NWH* in Jharkhand and Chhattisgarh were also the two largest values in the whole country. The smallest negative values of *Slope*\_*NWH* and *Slope*\_*NFHWH* were in Haryana. The *Slope*\_*NFHWH* in the mainland Gujarat had the second most negative value, while the value of the *Slope*\_*NWH* was the third-highest positive value.

**Keywords:** adaptive K-means algorithm; heavy industry heat sources; NPP-VIIRS; active fire data; night-time light data

#### **1. Introduction**

Over the few past decades, India has become one of the world's fastest-growing major economies and is now considered a newly industrialized country [1]. The amount of heavy industry, which is an important component of basic industry and provides technical equipment, power, and raw materials for all sectors of the national economy, has also soared in India [2]. This industry effectively supports the economic development of the country. However, this growth has been accompanied by a large increase in greenhouse gas emissions and other air pollutants from heavy industrial production [3]. Therefore, real-time maps of the layout of heavy industrial development are becoming important for studies of Indian economic development and air pollution issues [2,4].

Many scholars and nonprofit organizations or institutions have focused their attention on the global distribution of one or more energy types or industries. The British Petroleum (BP) company [5] and the International Energy Agency (IEA) [6] provide regular, annual reports of energy (coal, oil, gas, etc.) prospects. The Global Power Emissions Database (GPED) [7] was formed from individual power-generating units for 2010 [3]. In addition, the India Coal-Fired Power Plant Database (ICPD) [8] is also available for India. These databases include a large amount of information that can be used for mining and strategic development in India. However, traditional statistical methods usually involve a lot of human error; in addition, the real-time distribution of heavy industry in India is not available.

Satellite images, which can be considered to be objective, true data, have become the most effective way to monitor the dynamics of Land-Cover (LC) and Land-Use (LU) (also referred to as LULC) [9,10]. Heat sources, such as the combustion of fossil fuels in cement plants and steelworks and the flaring of petroleum gas in oil fields [2,11], are also vital for most heavy industries. Therefore, thermal anomaly products derived from remote sensing data provide new ways of revealing the objective and real-time distribution of heavy industry in India. Recently, it has been widely and well-used in the detection of global-scale self-ignition fire point data [12–16]. Also, the night-time thermal anomaly product from the National Polar-orbiting Operational Environmental Satellite System (NPOESS) Preparatory Project (NPP) Visible Infrared Imaging Radiometer (VIIRS) has been successfully applied in studies of volcanic activity [17] and oil exploitation [18]. NPP VIIRS night-time fire data (resolution 750 m) were used to identify industrial heat sources considering their time, space, and temperature information [11,19]. Also, better active global fire-points product named NPP VIIRS active fire product (VNP14IMG), with 375-m resolution and covering day- and night-time thermal anomaly, was provided by Schroeder et al. [20] and Giglio et al. [21]. It effectively provided an improved response for fires with small areas. Then, Ma et al. [2] proposed a heavy industry heat source detection model based on an improved adaptive K-means algorithm using long-term VNP14IMG data. This produced good results for mainland China; however, due to the complexity of the Indian geographical coverage, the precision was not so good when this was applied to India.

In addition, large and heavy equipment and facilities (such as heavy equipment, large machine tools and large buildings) are also important characteristics of heavy industry. So, the use of lighting is also common and necessary in those areas. Night-time light (NTL) data, especially the VIIRS day/night band (DNB) data, can provide the day and night distribution of lights for the whole world [22,23]. Therefore, in this study, NTL data were used to modify Ma's model [2]. The new heavy industry heat source detection model for revealing spatiotemporal patterns in and the development of heavy industry in India based on an improved adaptive K-means using VNP14IMG and NTL was then developed. As part of this study, VNP14IMG and NTL data were acquired and preprocessed. We adopted an improved adaptive K-means algorithm using long-term VNP14IMG data to construct heat-source objects. Then, many hot features, including geometric, statistical, and heat source attribute features, were extracted for each heat-source object. In addition, the initial heavy industry heat sources were discriminated from other heat-source objects using a threshold recognition model based on hot features. Finally, maximum night-time light data were used to delineate the final heavy industry heat sources.

The remainder of this article is organized as follows. Section 2 describes the study area, data sources, main data preprocessing steps, and methodology. Section 3 shows the experimental results that were obtained using the VNP14IMG and NTL data and discusses and assesses the distribution of heavy industrial heat sources in India. Conclusions are drawn in Section 4, and recommendations for future research are given.

#### **2. Materials and Methods**

#### *2.1. Study Area*

India is a country in South Asia, lying to the north of the equator between 6◦44 N and 35◦30 N and 68◦7 E and 97◦25 E. It is surrounded by the Indian Ocean, the Arabian Sea, and the Bay of Bengal. Since market-based economic reforms began in 1991, India has emerged as a global player with one of the fastest-growing major economies and is now considered a newly industrialized country [24]. It is also the world's second-most populous country (with more than 1.3 billion people) as well as being the most populous democracy in the world. India is a federal republic governed under a parliamentary system and comprises 29 states and seven union territories, giving a total of 36 entities (as shown in Figure 1). It should be noted, however, that Jammu and Kashmir state, marked by the red dashed line, lies within the disputed Kashmir region.

**Figure 1.** The 36 States and Union Territories of India.

*2.2. Data Sources*

#### 2.2.1. VIIRS Active Fire/hotspot Data

In this study, the VNP14IMG data were selected as input data for the evaluation of the distribution of heavy industrial heat sources in India. This product is based on reprocessed nominal-resolution Collection 1 data from the NASA Land Science Investigator Processing System (Land-SIPS) [20]. Using the MOD14/MYD14 algorithm, several modifications were implemented to accommodate the unique characteristics associated with the VIIRS 375-m data [25]. The newly improved 375-m data, compared to the traditional coarser-resolution (≥ v1 km) fire products, provide a greater response for fires that cover relatively small areas and improved mapping of large fire perimeters. So, it is well suited to support fire management as well as to meet other scientific applications' needs. VNP14IMG data (19 January 2012 to now) can be freely obtained from the Fire Information for Resource Management System (FIRMS) [26]. Three million nine hundred ninety-eight thousand four hundred sixty-five observed Indian fire hotspots, ranging from 19 January 2012 to 31 December 2018, were used in this paper, and their spatial density is shown in Figure 2.

**Figure 2.** Spatial density of the 3,998,465 fire hotspots in Indian regions (including Jammu and Kashmir state).

VIIRS Nightfire product (VNF), using Day/Night Band (DNB), near-infrared (M7 and M8), short-wave infrared (M10), and mid-wave infrared (M12 and M13) to detect subpixel heat sources, has been used in gas [27] and industrial heat sources detection [11]. So, VNF data were downloaded from the Earth's Observation Group (EOG) [26]. Their spatial distribution maps from VNF data and VNP14IMG data were made to compare in the study area (Figure 3) on 01/01/2018. It showed that VNP14IMG data were quite abundant in India. The fire/hotspot number of VNP14IMG was more tban five times than VNF. Its spatial distribution range was also bigger than the VNF data. So, VNP14IMG data were used lastly to detect heavy industries.

**Figure 3.** Spatial distribution comparison of VIIRS Nightfire product (VNF) and NPP VIIRS active fire product (VNP14IMG) on 01/01/2018. (**a**) The spatial distribution comparison of 301 VNF on 01/01/2018. (**b**) The spatial distribution comparison of the 1760 VNP14IMG hotspot.
