Onion is a high-value crop that is highly susceptible to price fluctuations in the Philippines due to weather conditions, global political situations, and other factors. Accurate mapping and yield monitoring is crucial for managing these fluctuations and ensuring stable supply chains. Unlike multispectral satellite images, hyperspectral images offer higher spectral resolution that enable it to differentiate subtle variations in the spectral signatures of onions compared to other crops. Thus, this study explores the effectiveness of PRISMA hyperspectral imagery for mapping onion fields through two distinct methodologies: K-means unsupervised classification and Linear Spectral Unmixing (LSU). The PRISMA image, captured on February 4, 2024, covers the area of Bongabon, Nueva Ecija, known as the onion capital of the Philippines, and its surrounding municipalities. The Level 2D product was denoised using Minimum Noise Fraction (MNF) by Forward MNF followed by Inverse MNF. The dimensionality of the image was then reduced using Principal Component Analysis (PCA). Three sets of data inputs - PC 1-2, PC 1-4, and 175 PRISMA bands - were classified using K-means. Separately, linear spectral unmixing was performed using four representative spectral signatures for each class - onion, rice, and soil – extracted from denoised PRISMA using known field locations. By comparing the outcomes of these methodologies, this research evaluates their accuracies in delineating the onions, with LSU providing more precise quantification of onion extent. The results highlight the potential of hyperspectral remote sensing in precision farming and in effective mapping and monitoring of onion yields to help mitigate market volatility.
KEYWORDS: RGB color model, Data modeling, Animal model studies, Algorithm development, Image processing, Image classification, Random forests, Modeling, Support vector machines, Sand
Escalating climate impacts prompt governments to act as seen in the fifth Conference of the Parties (COP), demanding eco-friendly practices to limit warming to 1.5°C. Carbon accounting is vital for global sustainability, requiring robust national monitoring of stocks and emissions. Remote sensing technology and satellite data enable modeling terrestrial carbon reserves, though challenges remain for coastal areas due to water attenuation. Ongoing studies aim to prove the technology’s viability, despite accuracy issues in capturing shallow coastal environments. With this being gap, this study developed a methodology to map a coastal environment using satellite data and machine learning. Sentinel-2 MSI, an open-source multispectral image, was utilized in this study. Geospatial derivatives such as ratios of the visible bands, bathymetry model using the Stumpf’s ratio and principal components which contained at least 90% of uncorrelated data were also integrated in the modeling process to improve benthic feature separability. Different combinations of the datasets were also explored in this study. Benthic habitat models were produced using Random Forest (RF) and Support Vector Machine (SVM) machine learning algorithms for each variable combination. The generated models generated overall accuracies ranging from 0.69 to 0.74 and 0.22 to 0.68 respectively. This translated to a maximum percent difference of 77% for the case of RGB model only and a minimum of 8% using all the variables. In terms of using different variable combinations, RF exhibited robust performance showing relatively consistent results compared to SVM which produced a wide range of accuracy values across the different models.
Water quality along the northern coast of Manila Bay is deteriorating due to anthropogenic influence, and the use of remote sensing is an effective tool for environmental monitoring. This study estimated the chlorophyll-a (chl-a) and total suspended matter (TSM) concentrations in Manila Bay from 2002 to 2016 and evaluated the possible environmental factors that contributed to the spatiotemporal changes in these two parameters. MODIS images were processed through the Case 2 Regional Coast Color model to determine monthly chl-a and TSM concentrations. Manila Bay was divided into six zones based on spectral characteristics. Each zone was then compared for the environmental variables precipitation, runoff, sea surface temperature, and wind speed downloaded from the ECMWF Reanalysis v5 global dataset. Zones 1-3 are located in the northern half of the bay and showed higher chl-a (3.2±0.9 to 8.3±2.2 μg/L) and TSM (2.0±0.7 to 11.0±2.5 g/m3 ) than Zones 4-6 (chl-a: 0.9±0.4 to 1.9±0.8 μg/L, TSM: 0.7±0.2 to 1.3±0.5 g/m3 ). The highest chl-a and TSM are in Zone 1, located at the mouth of Pampanga River, which is the largest watershed in Manila Bay. It is also an area with extensivemariculture activity. Within Zone 3 is the mouth of Pasig River, a localized area with anomalously high chl-a and TSM due to the high amount of organic load from urbanization. Pearson correlation of the environmental variables in each zone shows that precipitation (0.15-0.68) and runoff (0.38-0.79) are more correlated with water quality than sea surface temperature and wind speed. Paired t-test of chl-a and TSM also show a significant difference between the wet (June to November) and dry (December to May) seasons. Results suggest that water quality is largely influenced by precipitation and runoff. This means that effective river basin management could be the key to improving water quality in Manila Bay.
Seagrasses are distinct flowering plants which thrive underwater. They are part of a complex ecosystem that supports different forms of life. Recent studies found out that coastal wetlands – mangroves, saltmarshes, and seagrass, are far more proficient in sequestering and storing carbon than terrestrial ecosystems. Although seagrasses occupy only 0.2% of the area of the oceans, they sequester approximately 15% of total carbon storage in the ocean. Several remote sensing techniques are available to map and monitor seagrasses but most of them focus only on extent and area coverage. To estimate the carbon sequestration of seagrass beds, aside from extent, other parameters are needed such as leaf area index, percent cover, density, biomass etc., However, there are limits in mapping seagrass parameters using remote sensing. The reflectance measured by sensors is affected by other factors such as water absorption, turbidity, dissolved organic matter, depth and phytoplankton which affects the backscattering of energy. In this study, different remotely sensed datasets and field data were used to measure the parameters needed to estimate the carbon sequestration. Multispectral satellite images such as Sentinel-2 and PlanetScope were utilized to map the distribution and percent cover. High-resolution RGB images obtained by unmanned aerial vehicle (UAV) were also utilized to correlate field data gathered parameters. Field data such as species, percent cover, leaf area index, canopy height and above ground biomass were gathered in situ. Data extracted from different remote sensing technologies were put together to support the estimation of carbon sequestration of seagrass beds.
Water resource monitoring and management has been an important concern in the Philippines, considering that the country is archipelagic in nature and is exposed to a lot of disasters imposed by the global effects of climate change. The design and implementation of an effective management scheme relies heavily on accurate, complete, and updated water resource inventories, usually in the form of digital maps and geodatabases. With the aim of developing a detailed and comprehensive database of all water resources in the Philippines, the 3-year project “Development of the Philippine Hydrologic Dataset (PHD) for Watersheds from LiDAR Surveys” under the Phil-LiDAR 2 Program (National Resource Inventory), has been initiated by the University of the Philippines Diliman (UPD) and the Department of Science and Technology (DOST). Various workflows has already been developed to extract inland hydrologic features in the Philippines using accurate Light Detection and Ranging (LiDAR) Digital Terrain Models (DTMs) and LiDAR point cloud data obtained through other government-funded programs such as Disaster Risk and Exposure Assessment for Mitigation (DREAM) and Phil-LiDAR 1, supplemented with other remotely-sensed imageries and ancillary information from Local Government Units (LGUs) and National Government Agencies (NGAs). The methodologies implemented are mainly combinations of object-based image analysis, pixel-based image analysis, modeling, and field surveys. This paper presents the PHD project, the methodologies developed, and some sample outputs produced.
Proceedings Volume Editor (1)
This will count as one of your downloads.
You will have access to both the presentation and article (if available).
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.