Real-time forest fire monitoring in the Guanacaste Conservation Area: A LoRa-based sensor node system with satellite data integration

IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 1 of 16 IGLOC-2023, T, IP, x75424 Real-time forest fire monitoring in the Guanacaste Conservation Area: A LoRa-based sensor node system with satellite data integration Bryan Martínez-Aguilara*, Jason Martínez-Aguilara , Camila Barrios-Moralesa , Andrés Calderón-Quesadaa , Heillery Enríquez-Ramíreza , Samuel Medina-Pastranoa , Valeria Mesa-Alzatea , Daniel Picado-Rodrígueza , Cristian Vega-Romerob , Wilberth Corrales-Torresb , Francisco Salazar-Lópeza , Axel Batista-Murilloa , Braulio Soto-Alfaroa , Anthony Gómez-Ureñaa a Escuela de Ingeniería Mecánica, bryan.martinezaguila[email protected] , jason.ma[email protected].cr , camila.b[email protected] , jose.calderonqu[email protected] , heillery.enriqu[email protected] , samuel.med[email protected]c.cr , [email protected]c.cr , daniel.picadorod[email protected].cr , francisco.sala[email protected] , axel.ba[email protected] , brau[email protected] , [email protected] , Universidad de Costa Rica b Escuela de Ingeniería Eléctrica, cristian.vegaro[email protected] , wilberth.co[email protected].cr , Universidad de Costa Rica *Corresponding author Abstract The Guanacaste Conservation Area, declared as a World Heritage Site by UNESCO in 1999, covers approximately 2% of the country’s territory, and is host to 2.6% of the global terrestrial biodiversity, it is also mainly composed of dry forest. Despite recent efforts to recover from the damage caused by logging, poaching and misuse of lands, the forest remains at risk from fires mostly from anthropogenic origin, enhanced by extremely dry weather due to climate change, which requires significant effort from park rangers to detect and act on. This research outlines the development of a LoRa-based sensor node system for the detection and prevention of forest fires in the Guanacaste Conservation Area of Costa Rica. The proposed system developed in collaboration with the Universidad de Costa Rica, the Guanacaste Conservation Area and local communities, detects signals of potential fires through a sensor node system, by utilizing Bosch´s BME688 environmental gas sensor and addresses challenges such as scalability, durability, costeffectiveness and minimizing the environmental impact of the nodes. In order to withstand the extreme conditions of the environment, the system is equipped with a specialized enclosure that ensures that the gas sensors will get exposed to the ambient air whilst the sensitive parts of the electronics are protected from the environment. Additionally, satellite data was used in order to determine the optimal positioning of the sensor nodes within the conservation area according to fire-prone areas of the forest or proximity to other vulnerable objects. The system establishes correlations and detects gasses associated with combustion; whilst also minimizing the use of potentially harmful materials and elements commonly used in electronics. The proposed system also has a simple installation and operation process. The future plan of this system includes the possibility of enhancing satellite and climatic data, providing a more comprehensive understanding of conditions within the forest and aiding park rangers to improve response times to forest fire events. All while using satellite data to position the sensors optimally. Keywords: Forest fires, Sensor node system, LoRa technology, Environmental monitoring, Biodiversity conservation, Sustainable development. Nomenclature β: Optimum angle of inclination |φ|: Absolute value of the latitudinal location Acronyms/Abbreviations ABS: Acrylonitrile butadiene styrene ACG: Guanacaste Conservation Area (Área de Conservación Guanacaste) AI: Artificial Intelligence BSEC: Bosch Sensortec Environmental Cluster dBi: Decibel relative to isotrope dBm: Decibel milliWatts ICE: Costa Rican Institute of Electricity (Instituto Costarricense de Electricidad) IDE: Integrated Development Environment IMN: National Meteorological Institute (Instituto Metereológico Nacional) IoT: Internet of Things LoRa: Long Range LoRaWAN: Low Range Wide Area Network LPWAN: Low Power Wide Area Network NDVI: Normalized Difference Vegetation IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 2 of 16 PC: Polycarbonate PCB: Printed circuit board PLA: Polylactic acid POM: Polyoxymethylene PTFE: Poly(tetrafluoroethylene) PVC: Poly(vinyl chloride) SMA: SubMiniature version A SWR: Standing wave ratio WEEE: Waste Electrical and Electronic Equipment 1. Introduction 1.1 Objectives The main objective of this research is to design a sensor node based on a LoRa transceiver capable of aiding in the early detection of wildfires in the Guanacaste Conservation Area. In addition to that, finding optimal points of placements for those nodes utilizing satellite imagery, to improve the chances that these will detect possible fires as early as possible. 1.1.2 Space technologies and fire detection Satellite imaging technologies have gained increased importance in the past decade in various fields such as forecasting, infrastructure, navigation and others that might benefit from the information provided by the technology. In the field of forest fire prevention and response, the information made available is invaluable to scientists due to the ability to analyze hundreds of thousands of square kilometers of natural reserves. In this research, the aim is to use this technology to identify points of interest related to fires in the Guanacaste Conservation Area. Satellite imaging could also be used for the detection of wildfires, however, the time it might take a satellite to pass over a same area repeatedly or to pass over a specific desired area might be too long to be used on its own and therefore other tools such as sensor arrays can be used in complement to improve detection time when the area is not visible via satellite at the time. 1.1.3 Guanacaste Conservation Area and its significance The Guanacaste Conservation Area (ACG) is a protected wildlife area located in the northwest of Costa Rica. It is composed of Santa Rosa National Park, Guanacaste National Park, Rincón de la Vieja National Park, Horizontes Experimental Forest Research Station, and the Bahía Junquillal National Wildlife Refuge. It covers 163.000 hectares of land, representing 2% of the Costa Rican land surface area. Fig. 1 ACG location [1]. The ACG consists of five main tropical ecosystems: marine coast, dry forest, cloud forest, and rainforest, being the only conservation transect in the world [2]. It was declared a UNESCO World Heritage Site in 1999 because of its geographical, sociocultural, and biological significance, this to preserve its natural resources and protect genetic diversity for perpetuity [3]. Unfortunately, the ACG and other close private properties have reported 41 forest fires on 1426 hectares during 2022 [4]. Wildfires can have a significant impact on the local wildlife, by destroying or contaminating habitats, food, water, or shelters. Therefore, this initiative rises to find a way to help combat the serious impact generated by forest fires in the ACG. This research project has as its main purpose the early detection of forest fires so that they can be contained as quickly as possible and minimize the impact on the ecosystem. 1.1.4 LoRa as a telecommunication system for IoT devices. LoRa is a low-power, long range wireless telecommunications system which can be used for the implementation of IoT devices, having the LoRaWAN protocol as a control mechanism of a wide area network of devices, in our case the nodes, enabling many devices to communicate with a gateway using the LoRa modulation effectively [5]. 1.2 Background In 1986, the ACG was established, using as its foundation the 10,400 hectares of the Santa Rosa National Park (which was created in 1971), over the years this area has increased to its present-day size of 163,000 hectares. Its main aim is to preserve and restore the tropical dry forest and its surrounding ecosystems, such as cloud forests, rain forests, and marine/coastal areas. As a part of its conservation efforts, the ACG performs sustainable tourism practices so as to raise awareness and IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 3 of 16 familiarize people with the conservation efforts in the area. The ultimate objective is to ensure that the ecosystems will remain preserved for perpetuity. From the establishment of the ACG in 1986 until 2023, over 50 million USD has been invested in the development of the ACG as well as the conservation efforts through funding from nine different countries (including Costa Rica), over 50 international foundations, and more than 10,000 private donors [2]. This biogeographic region, covering roughly 2% of the country and 13% of Guanacaste province, contains about 335,000 species of terrestrial organisms, which accounts for around 2.6% of the global biodiversity. To put this into perspective, the number of different land species found in the ACG surpasses the total amount of such in Northern Mexico, the United States, and Canada combined [1]. Wildfires in forests can have multiple root causes, including natural ones such as lightning strikes, as well as manmade ones. There are two types of causes: structural, which means that it doesn’t start the fire, but it can contribute to it, such as the dry season in Guanacaste. The other cause is immediate, which means that it starts a fire, for example, lightning [6]. Worldwide, most wildfires are caused by humans (89% average from 2018 to 2022). Wildfires caused by lightning tend to be slightly larger and to burn more acreage (53% of the average acreage burned from 2018 to 2022) than human-caused fires [7]. In Guanacaste, 99.9% of the province's forest fires occur as a result of human action, whether it is caused by accident or vandalism [8]. The most common causes of wildfires in the ACG are: Human activity: As mentioned before, 99.9% of wildfires in Guanacaste are caused by human activity, such as campfires, burning of agricultural waste, discarded cigarettes, and fireworks. Additionally, intentional burning for land management purposes is common in the region. Drought: Drought conditions can increase the risk of wildfires, as dry vegetation is more susceptible to ignition. High temperatures: High temperatures increase the risk of wildfires as it dries out vegetation making it more flammable. In Guanacaste, temperatures can reach over 35°C [9]. Winds: Strong winds can help with the rapid spread of wildfires, making them difficult to contain and extinguish. In Guanacaste, strong winds are common, especially during the dry season, increasing the risk of wildfires [10]. Vegetation: Dry vegetation acts as fuel for fires, because it provides the material for the flames to burn. Invasive plant species, which includes Jaragua, can contribute to the risk of wildfires since they can spread rapidly [11], infiltrate into the local ecosystems and out compete with native plants therefore removing them from the ecosystem. Illegal hunting: Also known as poaching, can also contribute to forest fires in Guanacaste. Poachers often use fire to flush out animals from their hiding places, making them easier to hunt. However, this practice can quickly get out of control and result in wildfires that can devastate large areas of forest and wildlife habitats [11]. Hunting wild animals is illegal, whether it is commercial or sport hunting [12]. 1.3 Related Work The Dryad Silvanet Suite is a commercial wildfire detection system that includes a wildfire sensor module as part of the detection solution. This module utilizes the BME688 gas sensor, a supercapacitor energy storage element, a transmit power of 14dBm with LoRa telecommunication technology, LoRaWAN as the technology used to manage a possible network of these sensor modules, and a 60-minute data transmit interval [13]. This specific implementation, although similar in its sensor and the telecommunications technology that is proposed in the development for the implementation of a node in the ACG, does have a much lower transmit power than what is available in other standalone LoRa products, such as the Wio-E5 wireless module, which offer a 22dBm transmit power, which might be relevant for the implementation of sensor nodes over longer ranges due to possible area limitations on internet access for the implementation of data relaying stations necessary for the delivery of the node`s output data to the parks authorities. The 60-minute time frame between data transmits is also a consideration. In addition to this system, there are also existing research articles with the development of fire detection and warning devices as their focus, such as Wireless Sensor Network for Ignitions Detection: An IoT approach [14], which also utilizes LoRa as its telecommunications system, but utilizes a DHT11 sensor, which is limited to humidity and temperature measurements only, and utilizes a lithium-ion battery as its energy storage element. Another research article with the main objective of developing a fire detection system is Forest fire detection system using wireless sensor networks and machine learning [15], which implements a regression model to improve the accuracy of its wildfire identification capabilities. This approach utilizes multiple sensors in its design such as the DTH22, which can provide relative humidity and temperature data, and separate light and carbon-oxide sensors. It also uses the nrf24L01 transceiver module instead of LoRa based telecommunication, as well as lithium-ion batteries as its energy storage method. IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 4 of 16 2. Material and methods 2.1 Investigation on ACG In March 2023, a visit to ACG, the Santa Rosa National Park and Horizontes Experimental Forest Research Station was made, with the purpose of obtaining data about the zone such as the type of vegetation, data obtention at a controlled wildfire and a general visit in order to establish operational requirements of the node. Cellular reception was tested on various zones for the purposes of establishing the possible locations for the gateways, which will receive the data from the sensor nodes and relay it so that it can be processed and delivered to the park authorities. The ACG’s Nuevo Horizontes ranger station had cell phone reception, electric power and a high observation tower, all of which are ideal conditions to locate an initial gateway for the LoRa system. 2.2 Node sensor training chamber For the BME688 sensor to have the decision capabilities that allow it to detect fire, first, an algorithm must be built, and for this purpose a lot of data points are needed. The BME688 Development Kit needs to be tested in a wildfire simulator to prove that the sensor works, collects data, and to assess the sensor´s performance. The smoke needs to be contained in an environment because the sensor needs to be exposed to smoke for a long period to collect the most amount of data possible, while using up small amounts of fuel. To make this possible, a testing chamber was built to contain smoke in a controlled environment. The testing chamber consisted of a steel water tank cut in half with a grinder so that the machine could lay horizontally on the floor. A door was made to introduce the material that is to be burned. Additionally, at the top of the tank, a steel pipe was welded vertically, and a steel hook was welded to support the sensor that was going to be placed inside. The gasses that are being burned in the floor bottom of the tank will rise and come out through the pipe, where the node will also be placed and ready to detect the gasses. Fig. 2 Testing and training chamber. The BME688 Development Kit was installed on the hook at the top of the cylinder. In order to collect data first the sensor was left inside the chamber without any smoke or fire for a minimum of 30 minutes. Then, a small fire was started, using either plant material or candles as fuel. This data is collected in the BME688 development kit and is then used for further analysis. 2.3 Node´s protective case To protect and secure the device, a protective case was designed. It has a rectangular box shape that includes an energy storage element and a printed circuit board inside. The case has a lid on its top where the solar panel is located and delivers energy to the power management system during sunny conditions. Due to the environmental conditions in the ACG and the susceptibility of the electronics required to detect and transmit data, being the supercapacitor the most critical component that must not exceed a temperature of 70 °C a protective case was designed (Figure 2) to ensure its survivability. The case structure is composed of three major elements: box, cover and ventilation ducts. Box: this piece is the main structure to which all other elements of the design are attached. Inside there are 4 standoffs into which the electronic board is screwed. The design provides a clearance of at least 1 cm between the electronics and any of the internal walls of the box, this feature is intended to minimize heat transfer by conduction from the outside of the encapsulation to the electronic components most susceptible to high temperatures. Another feature is the slots located at the sides of the bottom part of the box, this allows the assembly to be easily strapped to a tree with flat rope. Cover: provides support to the solar panel and covers the upper part of the box. It prevents the passage of rain into the interior of the case. A gap is presented between the bottom of the box and the solar panel to decrease the heat transfer between these two elements. Ventilation ducts: pieces attached to two of the outer walls of the box. They are designed to prevent the passage of water into the box in windy rain scenarios. Ventilation ducts also ensure the passage of air currents through the interior of the enclosure for effective detection of wildfire smoke. IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 5 of 16 Fig. 3 Render of the assembly of the node protective case with the solar panel and antenna. 2.4 Node´s electronics In the design of the node´s electronics, components were selected in order to meet a goal of using a solar panel with less than a 1 W power output capacity, this due to size limitations. Said components include a microcontroller, a sensor with the ability of detecting the fires, a LoRaWAN capable radio and energy management system. The low power requirement of each component was of utmost importance as it directly influenced the power management system. The decision process for each component is critical in determining the final performance of the sensor node. The sensor was the most crucial component in determining the other components of the node´s electronics since its performance is essential for the accurate detection of fires, and its power consumption determined the rest of the components needed for the power management system. The BME688 gas sensor from Bosch was chosen based on its ease of use and sensitivity on tests. The sensor´s active power consumption is around 3.9mA on scanning mode and 1µA while on sleep mode, making it an ideal choice for the node´s low power requirements. Bosch provides the BME AI-Studio software and the BSEC 2 library, which enable this sensor to use artificial intelligence in the decision-making process which determines whether the scanned air sample contains gasses associated with a fire, additionally, the AI model is easy to train since Bosch provides the software needed to train the sensors. This approach does however carry a disadvantage, since Bosch is restrictive of the library by making it closed source, and only giving out a handful of compiled binaries needed to run the AI algorithm on specific microcontrollers. Only microcontrollers whose cores are listed on Table 1 [16] can take advantage of the sensor´s AI capabilities. Table 1. Microcontroller cores supported by the BSEC library [16]. Microcontroller platform Core type Cortex ARM M0, M0+, M3, M4, M4 FPU, M7 AVR 8 bit MegaAVR, XMEGA AVR 32 bit 32-Bit AVR UC3 ESP8266 ESP8266 ESP32 ESP32 MSP430 MSP430 For the microcontroller, the SAMD21 microcontroller was chosen, due to its low power consumption relative to the planned energy goal. This microcontroller is fully compatible with the Arduino IDE, providing ease of use for developers. In addition, this microcontroller can go into a sleep mode that lowers power consumption to just 15µA, making it an appropriate choice for this application. Another important factor in choosing this microcontroller is that it is built using Arm m0+ cores, which are fully compatible with Bosch’s BSEC. In addition, the choice was made to opt for a development board that already had the ATSAMD21G18 microcontroller integrated. The Seeeduino XIAO, developed by Seeed, was selected for this purpose since it included the necessary components needed for the SAMD21 to operate, while adding minimal cost, and allowing for a faster development time since there was no need to place and calculate all the supporting components for a microcontroller on the proposed PCB. To power the previously mentioned components, a system consisting of a solar panel, a solar energy harvesting controller, an energy storage element and a switching DC to DC converter was proposed. Once this was defined, the power management system was to be based around a lithium-ion supercapacitor energy storage element. The decision for this was based on safety and durability under the conditions in which the system is supposed to function under. Lithium-ion batteries, despite having a generally much higher energy density, much lower cost per Wh of energy storage capacity, as well as a vast amount of resources for the implementation in such a system, have undesirable characteristics regarding safety, the most relevant being their susceptibility to suffer violent exothermic reactions when put under excessive stress conditions, such as high temperatures, mechanical damage such as puncture or crushing, as well as electrical abuse such as short circuits, or charging the element past the voltage specification [17]. IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 6 of 16 Due to the conditions in which the node is to be implemented, which include the consideration of potential malicious handling of these devices, such as intentional destruction, the temperature conditions to which the device might be exposed to, and possible failures in the electrical components which might cause over voltage or short circuits, the use of lithium-ion supercapacitors is preferable due to their significantly less violent response to abuse in all of the forms mentioned previously, which, although will cause irreversible damage to the energy storage element, such as swelling, capacitance degradation, and even gas leaks due to oxidation of the internal materials, are not as likely to suffer violent exothermic reactions [18, 19], while retaining a similar ease of implementation due to the lower but similar voltage working range that allows the use of electrical components designed for lithium-ion batteries with slight modifications in this implementation. Lithium-ion capacitors also have a much higher cycle life than lithium batteries, which is relevant for this system due to a desired long service life and low maintenance requirements. The charge controller selected is the E-peas AEM10941, which can manage both the extraction of energy from a solar panel, including measures for energy harvesting under unfavorable conditions, and charging of the energy storage element, with additional safety features such as over charge and over discharge protection of the storage element [45]. The reason why this specific controller was chosen over other charge controllers with similar performance characteristics such as the BQ25570 from Texas Instruments or the SPV1050 from STMicroelectronics is both its low external component requirement amount required for full functionality as well as the ability for it to be fully configured with only external resistors, which makes the manufacture of the Power management System easier due to not having to program these controllers via a serial communication protocol individually, as well as a lower price compared to these. To generate a stable 3.3V output a Nisshinbo Micro Devices RP605Z334A was selected, this due to its 0.3µA operating quiescent current, which means the amount of energy it consumes to operate is negligible compared to the devices it power, such as the 15µA the processor selected consumes under sleep mode, as an example. It also provides an efficiency of at least 75% at the lowest voltage provided from the energy storage element before shutting down, while being above 85% for most of the range of operation of the energy storage element, this according to the device datasheet [44]. 2.5 Sample collection Fig.4 Proposed sensor node system overview Fig.5 Prototype Node with components Data collection for the Bosch BME688 gas sensor was done utilizing the Bosch BME688 Development Kit, which carries 8 BME688 sensors on board, the microcontroller chosen for the development kit was the ESP32 Feather from Adafruit since Bosch provides a flashing utility that makes this microcontroller directly compatible with their BME AI Studio software. 2.6 Antenna performance data collection In order to measure the performance of antennas, the NanoVNA spectrum analyzer was used, in addition, NanoVNASaver was utilized to extract the data from the NanoVNA and save it into an image. 2.7 Definition of possible locations for the nodes To define the different location points of these nodes, the main consideration was given to the required range of operation in areas where the probability of a forest fire is higher. Based on the information obtained, a prioritized IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 7 of 16 order was generated for the installation of the devices. Valuable insights were obtained using satellite imagery, which was utilized as a crucial tool to identify the most strategic points. To achieve this, the proposal is to use satellite imagery retrieved from the Copernicus SENTINEL-2 satellite, combined with QGIS software, in order to identify causal factors that could be correlated with forest fires, such as areas where previous burns have occurred, water scarcity, vegetation in different areas, proximity to population centers, and other factors that may not have been previously considered. During the analysis of satellite images, a particular tool called Normalized Difference Vegetation Index (NDVI) was used. This tool provides us with a scaled view of the amount of vegetation in a particular area. Taking this and the proximity of surrounding livestock farms into account, possible locations for the nodes were selected. 2.8. Enhancing satellite data and firefighting on the ACG Satellite imagery can function as a tool for wildfire detection. However, this detection mechanism has a major flaw: due to the nature of satellites, they are constantly orbiting Earth and are not able to always maintain a fixed position. Therefore, they may not be able to detect a wildfire immediately if it occurs in an area outside the scope of their monitoring during their orbit, so by the time they detect the wildfire, it may already be too late. This is where the node system becomes crucial. Since nodes are always situated within a fixed area, they are capable of detecting wildfire in its early stages. This detection system complements and enhances the performance of satellite imagery in this area, resulting in a much more optimal and faster wildfire detection system. Therefore, the use of nodes as a complement to satellite imagery is of utmost importance to improve the early detection of wildfires. On the other hand, the installation of nodes represents a significant advantage at night. Even though there are different towers strategically distributed throughout the Guanacaste Conservation Area (ACG) for the purpose of identifying if there is a fire starting in an area, so that it can be extinguished as soon as possible. The successful functioning of these towers lies on the identified smoke generated by the fire, which depending on its size, can be seen from kilometers away. However, if a fire occurs during nighttime, it is impossible for park rangers to identify such fires with the naked eye. With the use of these nodes, it is no longer necessary to visually identify the fire, as the sensors installed in the nodes can detect it. This optimization would allow the identification of a fire in its early stages, making the process of extinguishing it much more efficient. 2.9. Determination of the internal temperature of a node´s protective case prototype under direct sunlight conditions Due to cost, complexity and time constraints, a different material and additive manufacturing technology were implemented to build a prototype of the protective enclosure. The building material was PLA, with 15% of infill. An experiment was carried out to measure the dry bulb temperature values inside a prototype of the protective enclosure of the node. It consisted of placing the enclosure under direct sunlight for 2 hours, to ensure stable thermal conditions. A BKP60 banana plug type-K thermocouple was positioned in the center of the interior of the protective casing, measuring the internal air temperature, while another thermocouple collected data on the ambient temperature. 3. Theory and calculation 3.1 Photovoltaic panel One of the main factors influencing the choice of the type of solar panel was the peak power rating, which should be sufficient to meet the consumption demands of the system and at the same time provide sufficient additional energy to charge the supercapacitors, so that the node can remain powered on during the night and during adverse weather conditions. This parameter was determined from the average consumption of the circuit over a period of approximately 10 minutes cycle with three phases of variable consumption, of which approximately nine minutes correspond to a sleep period, one minute of data collection and finally about two seconds of data transmission through the LoRaWAN network protocol as shown in the following table. Table 2. Node power consumption in mA/h. Mode Consumption (mA/h) Sleep 1 Transmission 126.3 Collecting 22 Average 3.52 From the data showed in Table 2, and, considering that the operating voltage of the node’s components is 3.3 V, it was determined that the energy management system must provide an average of 0.0116 W to power the node´s electronics, and with a limitation of 110mA from the solar harvesting controller [45], the maximum charge power deliverable to the energy storage element is 4.2wh at the 3.8v rated maximum voltage of lithium-ion supercapacitors. Due to this a panel able to deliver at least 0.55 W of energy was set as the baseline to account for a 30% positive tolerance, which accounts for the performance degradation of the selected solar panel type. IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 8 of 16 With this the system will charge two 250 F supercapacitors, totaling 500 F of capacitance, with a working voltage range that goes from 2.5v up to 3.8v, resulting in a 0.55 Wh storage capacity. This should be able to sustain the system without charging for up to 40 hours with a raised efficiency of at least 75% from the dc-to-dc converter circuit based on the RP605Z334A [44]. Due to the nature of the project, the chosen panel is of the monocrystalline silicon type, because it presents a good quality-price ratio, also, this type of material features an efficiency in the range between 19% and 26% in transforming solar energy into electrical energy. Within the panels considered the Voltaic brand stands out, which have some models that have an efficiency of 22% or more, on the other hand, considering the environmental conditions, another viable option is the polycrystalline panel because of its greater resistance under conditions of high solar incidence and high temperatures. Theoretically, monocrystalline silicon cells present an efficiency of 26.7% versus 21.9% efficiency of polycrystal<lline silicon panels [20]. However, since it was determined that the panel should be as small as possible to be less visible to poachers and people who could damage them, even in order to interfere as little as possible in the ecosystem, polycrystalline solar panels were discarded. Another convenient factor to analyze was the solar radiation potential for the production of photovoltaic energy in the province of Guanacaste (Figure 1), corresponding to an annual average of between 17 and 20 MJ / m² per day according to data obtained from the Costa Rican Electricity Institute (ICE for its acronym in Spanish), being one of the areas that receives more irradiation in the country [21], also according to determinations of the National Meteorological Institute (IMN) the annual solar brightness that receives the north pacific corresponds to 7 hours or more [22], and as shown in Figure 2 and Table 2, the annual average of 7.5 hours for the Santa Cruz region will be considered as the number of effective hours that can be used for the calculations of the daily production capacity of the photovoltaic panels, at least for the dry season, also, taking into account the lower solar brightness values that occur during the rainy season, the duty cycle of the node can be configured to ranges that allow greater energy savings to compensate for the decrease in photovoltaic energy, contemplate that the incidence of forest fires is considerably reduced during this period. In this sense, Sebastián Romero (2016) points out that due to the latitudinal location and the small size of Costa Rica and analyzing it in conjunction with the constant of angular motion in Costa Rica the sun has a trajectory of approximately 180°, whose variations at different times of the year are practically negligible as is the length of the day [23]. Thus, taking an average value will be very close (with some climatological exceptions) to the minimum value that the region can receive during the dry season and the rainy season. Fig. 6 Monthly average daily global solar radiation. Taken from the Costa Rican Institute of Electricity [21]. Fig. 7 Average sunshine in the North Pacific region. Taken from the National Meteorological Institute [22]. Table 3. Average daily hours by region in Guanacaste. Taken from the National Meteorological Institute [22]. On the other hand, the use of solar energy also depends on the thickness of the clouds and the amount of sky cover, atmospheric transparency, latitude and elevation of the site, Guanacaste most of the time in the summer season presents clear skies and high solar incidence throughout the all-day reasons why the values of radiation potential and solar brightness are higher than in other regions of the country. The role played by cloudiness in the Guanacaste area is variable throughout IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 9 of 16 the year and there is little information about it, however, according to the ICE this deficiency can be compensated with the data of solar brightness, and even the global radiation potential, that includes direct radiation and diffuse radiation that reaches the different areas of the country, The diffuse component varies from 20% on a clear day to 100% on a completely cloudy day, the thickness of the clouds affects the amount of reflected and blocked radiation [21], therefore the energy generation depends on the duration of cloudiness throughout the day as well as its thickness. Regarding the altitude, this is around 50 meters above sea level for the Santa Cruz area, the IMN states that in the low areas of the country the air generally contains more humidity and suspended particles than the high areas [22], however, despite having a low altitude, the hours of solar brightness compensate for this deficit. To reduce energy consumption to a minimum, the use of solar trackers is discarded; instead, the angle of inclination capable of generating a greater use of radiation in the summer season is determined. According to Norma Flores and Miguel Domínguez, the solar panels should be placed in such a way that throughout the year they are able to take maximum advantage of the available solar radiation; the ideal orientation is towards the geographic south in the northern hemisphere. Deviations of up to 20° to the southeast or southwest with respect to the orientation do not significantly affect the performance of the system. To maximize utilization, it is necessary to decrease latitude by 15° [24]. On the other hand, Oscar Perpiñán (2013) proposes the following formula for the ideal inclination calculation [25]: β = 3,7 + 0,69 |φ| (1) According to data taken from Google Earth, the Guanacaste Conservation Area (ACG by its Spanish acronym) Santa Rosa sector is located 10°50'16'N 85°37'01'W. Therefore, applying the formula used by Flores and Domínguez gives an angle of inclination of 5° below the horizontal, while using the formula proposed by Perpiñán the optimum angle of adjustment is 10.6° with respect to the horizontal position. Therefore, it is concluded that the optimum angle will present a degree of inclination very close to zero due to the relative proximity of the country with respect to the equator. Another advantage of the use of solar panels is the minimal environmental impact, since most of the components that make them up do not represent a risk to the environment, in addition to being recyclable; the greatest environmental burden is associated with the extraction and transportation processes. In general, the number of materials required is small, due to their scale and use. At the same time, it does not generate toxic products or gas emissions associated with its implementation. Photovoltaic energy even helps to reduce environmental problems such as the greenhouse effect due to the reduction of CO2 emissions, acid rain caused by sulfate emissions and photochemical fog caused by nitrate emissions. Regarding the environmental impact, Sierra, Vásquez and Ramírez (2021) point out that photovoltaic panels are a renewable energy that is mitigating the production of greenhouse gasses, mainly carbon dioxide CO2 [26], which is noteworthy because they are a source of natural energy and that it is sustainable for the environment, on the other hand these authors point out that the materials implemented in the panels are materials of easy access and recovery, these being glass, aluminum sheets and polymers, which are classified as special handling waste type WEEE (Waste Electrical and Electronic Equipment), recovering 96% of material in polycrystalline panels and 98% of material in thin film panels, generally reaching 90% recovery and reuse of waste components for new industrial processes [26]. 3.2 Node’s Case Material Selection A major coverage of the park with nodes is an important part of the project, each node will have its own case, which will provide protection from environmental and human provoked damage. The cases must be easily and quickly manufactured and replaced in case it presents any imperfection or critical damage. One way of achieving this is with mass production methods, for this project it is intended to use 3D printing technology for the cases prototypes and subsequently proposing a plastic injection process for mass production in the future. Different plastics are used for injection molding, some of the most common are the nylon 6/6, polyoxymethylene (POM), polystyrene, acrylonitrile butadiene styrene (ABS), polytetrafluoroethylene (PTFE), polyvinyl chloride (PVC) and polycarbonate (PC). These plastics will be compared and analyzed to help choose the one that best adapts to the node's case needs. One of the most important properties to consider for the material selection is the service temperature, which is going to help the material to stay in the wild forest without losing their physical properties. For this analysis a lot of technical data sheets from different companies that distribute these materials were collected, with the purpose of getting an estimate of the actual characteristics that could be obtained when using each plastic. IAF Global Space Conference on Climate Change (GLOC 2023) - Oslo, Norway, 23-25 May 2023. Copyright ©2023 by the International Astronautical Federation (IAF). All rights reserved. GLOC-2023-F1.2.3 Page 16 of 16 [31] Elaplast, PROPIEDADES FISICAS POLICARBONATO (PC), N.d, https://www.elaplas.es/wpcontent/uploads/policarbonato.pdf, (accessed 03.24.23). [32] Polylanema, Plasticos de Uso General: PVC, N.d, https://www.polylanema.pt/client/files/0000000001/pvc -es_2046.pdf, (accessed 03.24.23). [33] Ensinger, TECARAN ABS grey, N.d, https://www.ensingerplastics.com/eses/semielaborados/plastico/tecaran-abs-grey#/producttechnical-detail-collapse-item-1-lvl-1, (accessed 03.21.23). [34] RUNSOM, Plásticos resistentes a los rayos UV: diferencia entre polipropileno y nylon, N.d, https://www.runsom.com/es/blog/plasticos-resistentesuv-diferencia-entre-polipropileno-y-nylon/, (accessed 03.23.23). [35] Mexpolimeros, POM (polioximetileno) - Resinas Acetalicas, N.d, https://www.mexpolimeros.com/acetal.html, (accessed 03.20.23). [36] Cole-Parmer, UV Properties of Plastics: Transmission and Resistance, N.d, https://www.coleparmer.com/tech-article/uv-propertiesof-plastics?tlg=es-ES, (accessed 03.24.23). [37] Standard Fluoromers, Standard Specifications for PTFE and Filled PTFE, N.d, https://www.standardptfe.com/standard-specifications-for-ptfe-and-filledptfe.php, (accessed 03.23.23). [38] A. Isan, ¿Por qué el PVC es tóxico y contaminante?, Ecología verde, 22 November 2017, https://www.ecologiaverde.com/por-que-el-pvc-estoxico-y-contaminante494.html#:~:text=Cuando%20se%20produce%20o%20s e,inmunol%C3%B3gico%20y%20el%20sistema%20rep roductivo.%22, (accessed 03.24.23). [39] J. Mark, Polymer data handbook, Oxford University, Press, 1999. (363-367). [40] Ensinger. TECANAT natural, N.d, https://www.ensingerplastics.com/esbr/semielaborados/plastico/policarbonato-tecanatnatural#/product-technical-detail-collapse-item-1-lvl-1, (accessed 03.21.23). [41] National Oceanic and Atmospheric Administration, Discussion on Humidity, N.d, https://www.weather.gov/lmk/humidity#:~:text=Warm %20air%20can%20possess%20more,if%20the%20air% 20is%20warmer, (accessed 05.12.23). [42] Bosch, Gas Sensor BME688, 2022, https://www.boschsensortec.com/products/environmental-sensors/gassensors/bme688/, (accessed 04.30.23). [43] CB WORLD, what is SWR and why is it SO important? N.d, https://www.wearecb.com/what-isswr.html, (accessed 05.12.23). [44] Nisshinbo Micro Devices, RP605x Series, N.d, https://www.nisshinbomicrodevices.co.jp/en/pdf/datasheet/rp605-ea.pdf, (accessed 05.10.23). [45] e-peas semiconductors, AEM1094, N.d, https://epeas.com/wp-content/uploads/2022/09/e-peasAEM10941-datasheet-solar-energy-harvesting.pdf, (accessed 05.10.23). =======================================