We studied a 2.4m long coral Porites core (U-Th dated 8986±28 to 7870±29 yr BP) collected from Agongdian Coral Reef limestone, southwestern Taiwan to reconstruct the paleoenvironment. Based on the X-ray photography, carbonate powder was drilled along the maximum growth direction for stable carbon and oxygen isotope and Sr/Ca ratio analyses.Oxygen isotope values of coral Porites range from -3.4‰ to -6.6‰ with a mean oxygen isotope value of -5.2±0.7‰ (mean±1s; N = 1634). Based on the fluctuation of oxygen isotope values, 106 oscillations with amplitudes between 0.5‰ and 1.0‰ were observed. Carbon isotope values of coral Porites are between -4.8‰ and 1.0‰ (-1.4±0.8‰). Sr/Ca ratios are between 8.3 mmol/mol and 9.8 mmol/mol (9.1±0.4 mmol/mol; N = 237).Nine temperature stages can be recognized bared on Sr/Ca ratio estimated mean temperatures. In ascending order, these mean temperatures are approximately 32.3°C (N=8; 8700 yr BP), 22.3°C (N=22; 8640 yr BP), 25.0°C (N=30; 8595 yr BP), 21.3°C (N=40; 8525 yr BP), 24.4°C (N=19; 8400 yr BP), 20.9°C (N=58; 8350 yr BP), 30.0°C (N=20; 8220 yr BP), 21.4°C (N=17; 8200 yr BP), and 29.6°C (N=23; 8090 yr BP).Adopting the Sr/Ca temperatures and using the equation of Shen et al., (1996), the calculated Early Holocene oxygen isotope values of seawater were -0.3‰ (N=71) for summer and 0.4‰ (N=44) for winter in SW Taiwan (ranging from -2.0‰ to 2.0‰). The calculated early Holocene seawater oxygen isotope values were either greater than or comparable to the modern SW Taiwan seawater oxygen isotope values observed. Thus, strength of SW monsoon/precipitation may have fluctuated in SW Taiwan during this period of time.

The early Holocene (~10 – 6 kiloyears before present, ka BP) was characterised by intensified monsoon, which was also punctuated by an abrupt weakening event at ~8.2 ka BP. During this period, the northern hemisphere insolation was at its maximum and atmospheric temperature were higher than present. Investigating the climate variability during the early Holocene can therefore be serve as a benchmark for future climate projections. Singapore lies within the Indo-Pacific Warm Pool (IPWP) region, that drives the global hydrological cycle and is affected by complex hydrological processes such as the north-south migration of the Intertropical Convergence Zone (ITCZ), Indian Ocean Dipole and El Niño-Southern Oscillation. We used stable isotope analysis on selected benthic foraminifera shells from a sediment core of the Kallang River Basin, Singapore, to understand the drivers of rainfall variability during the early Holocene and how global climate events such as the ‘8.2 ka event’ are expressed in this region.

Project-based Learning has gained traction as an alternative means of engaging the learning of students. With the advances and accessibility of techonology tools that were once the enclaves of only researchers, citizen science has taken a big step forward. Students would find it possible to engage in research activities with a very low budget that they can afford. The author will share some of the projects that were undertaken by the students in the field of Geo Sciences. In this presentation, the author will also share with the audience the AOGS International Geoscience Challenge competition to encourage high school students to participate in. 

In Taiwan, large earthquakes often induce severe disasters and lose of human lives and money. Such as 1999 Chi-Chi earthquake occurred in central Taiwan, thousand of buildings collapsed increased the loss lives. It remind us the importance for the earthquake resistant structure for the buildings in our living environment, especially for the campus. In order to better explanation for the effect of these structure applied for the buildings to K-12 students, the courses and teachable aids are investigate and design using the experiments of the combination of the architecture devices made by wood and magnet. By the comparison before and after equipped the earthquake resistant structure devices, students can easily observe the stable contribution for these earthquake resistant devices. In this study, we introduces the investigating process for the teachable aid and the course design for the class and community. Finally, we share the practical experiments in the classroom and the museum and study the learning performance for K-12 students.

The Earth’s ionosphere can reflect the radio wave to assistant signal propagating to far distance, but also affect radio propagation by its electron density and structure. The ionosphere not only influences the radio wave from one ground stations to the other, but also communications between satellites and ground sites. Therefore, satellite data transmission, video and audio broadcasting, navigation and positioning, and orbit control are all affected. The FORMOSAT-5 was launched with the advanced ionosphere probe (AIP), which could provide precise and accurate measurement for ionospheric density structure. On the other hand, the FORMOSAT-7 using radio occultation technique to measure the Earth’s atmosphere and ionosphere which could provide most accurate data globally. Therefore, many course modules including lectures, paper models and hands-on experiments, were designed and conducted by professors and researchers in Nation Central University to introduce knowledge on radio communication, space science and space weather. There courses are performed to for elementary and junior high school students, and general public in Taiwan in past two years.

Proliferation of high-resolution remote sensing sensors along with the recent advancements in artificial intelligent attract much attention of remote sensing community for the development in technologies for rapid monitoring of environmental and climate events in larger scale during last decade. Such potentials in remote sensing applications are crucial in timely detection and management of frequently occurring natural disasters such as floods to minimize the loss of human lives and damage to the economy. Moreover, most of the near real-time flood mapping applications did not pay much attention on large scale inundation depth estimations even though it is an extremely vital measure for rescue operations as well as further improvements in infrastructure developments. Therefore, this study proposes a methodology to rapidly map inundation extents along with inundation depths from high resolution satellite images using convolutional neural networks. The networks were trained and tested using ~2600 km² of SPOT-6 and SPOT-7 satellite images (1.5 m resolution) observed during several flood events occurred in Japan from year 2015-2019. Obtained results revealed that the networks were achieved very competitive accuracy in mapping flood extends. Moreover, it has been found that the proposed inundation depth estimation algorithm is capable of estimating inundation depths relatively well with a speed of 0.4 s/km² for the tested data set.

Although most of smart watches are not made for medical use, they become popular for the function of health monitoring such as heart rate (HR). The main objective of this study is to evaluate the effect of fine particulate matter (PM_2.5) on HR using the low-cost sensing (LCS) devices and smart watches in commuters. In this study, we recruited 50 subjects living in an urban community in the southern Taiwan. Subjects were asked to wear a LCS device, i.e., AS-LUNG-P (Academia Sinica-Lung-Portable version, integrated by our team) and a smart watch (Forerunner 35, Garmin, Garching, Germany) for 7 consecutive days. AS-LUNG_P was used to monitor personal PM_2.5 levels, temperatures, and 3-axis accelerations. Subjects were also asked to completetime-activity diaries (TADs) at 30-min intervals to determine the duration when subjects spent time commuting. The commute mode with indoor (i.e., sedans, taxis, buses, mass rapid transit (MRT), and trains) and outdoor (i.e., walking, bikes, and scooters) environments were evaluated separately. The generalized additive mixed model (GAMM) was used to evaluate the association between 5-min average PM_2.5 concentration and HR adjusting age, gender, body mass index (BMI), activity, temperature, and time of day. In terms of commute mode with indoor environments, results indicated that an increase in PM_2.5 concentrations was significantly associated with an increase in HR. However, we found that an increase in PM_2.5 concentrations when subjects commuted in outdoor environments was significantly associated with a decrease in HR. This is probably because diffusion of PM_2.5 in outdoor environments. This study demonstrated the application of smart watches without medical certifications to assess the association between PM_2.5 exposure and HR. In future research, the HR results and its relationship with PM_2.5 of smart watches should compare to that of devices with medical certifications.

The importance of developing a low-cost and non-invasive alternative methodology for liquefaction assessment is emphasized as significant portions of land area of Greater Metro Manila is exposed to such hazard. Geomorphic setting ideal for liquefaction such as beach zones, sandbars, and marshes comprise a considerable fraction of the Metropolitan. Moreover, its proximity to the active Valley Fault System increases the risk for potentially damaging and liquefaction-causing ground shaking events. As traditional methods for liquefaction assessment, such as Standard Penetration Tests (SPT), are typically costly and labor intensive, the benefit of geophysics as alternative in liquefaction assessment is further highlighted. This study employs three geophysical instruments, namely (1) Ground Penetrating Radar (GPR), (2) Refraction Microtremor (ReMi), and (3) Three-component Microtremor tested in conjunction with a portable drilling equipment named Screw Driving Sounding (SDS). Deployment of these equipment is faster, generally cheaper, and leaves minimal footprint afterwards. To date, forty-three (43) schools out of the targeted sixty (60) have been tested using the said geophysical instruments and Screw Driving Sounding machine in Metro Manila and Bulacan Areas. Liquefaction Potential indices for each school was calculated. Out of the 43 schools, 34 schools also have available SPT data that were compared with the obtained geophysical data. Results from the surveys and the comparison to SPT results suggest that this new methodology can be an effective and non-destructive technique for shallow subsurface characterization useful for quantifying liquefaction potential.

Seismic liquefaction occurs when strong ground shaking propagates through soil layers causing loosely-packed surface materials to increase pore pressure, distort sediment structure, and eventually lose material strength triggering settlement. This process will transform sediments in solid-state to behave in a liquid state, hence, rendering structures built on top of liquefiable ground to sustain considerable damage. Conventional methods of quantifying liquefaction hazards involve in-situ drilling techniques such as the Standard Penetration Test (SPT) or the Cone Penetration Test (CPT). These methods are the standard in current engineering practice. However, they are invasive and destructive on the sites and costly in terms of labor, time, and money.
In this study, the Horizontal-to-Vertical Spectral Ratio (HVSR) of microtremors was used. It is a passive seismic geophysical technique, and we aim to evaluate its effectiveness in assessing liquefaction hazards. An array of single-station measurements were performed within the confines of selected sixty (60) experimental school sites in the coastal lowlands of the Greater Metro Manila Area (GMMA), Philippines. Measured predominant periods and their relative amplitudes were correlated with calculated liquefaction potential indices (LPI) from available borehole data. Moreover, the LPI values were compared to the seismic vulnerability index (Kg) of Nakamura, a qualitative index that estimates deformation level due to local site effects. This study suggests that the employed methodology can be a quick, non-invasive, and cost-effective complement to existing borehole measurements in the assessment of liquefaction. Such a method can be adapted between borehole data gaps to extrapolate information where downhole data is limited or not always available.

Earthquakes are among the most feared and destructive of all natural disasters. China is one of the most earthquake-prone countries in the world. However, it is difficult to predict the entire number and specific location of earthquake casualties, which is the critical issue resulting in lagging of earthquake emergency management. Rapid judgment of the trapped personnel location is the important basis to identify the emergency supply demands and carry out the search and rescue work after the earthquake. Through analyzing the main influencing factors, we constructed an empirical assessment model of people trapped in collapsed buildings caused by the earthquakes based on the km grid data and LBS technology. The accuracy of the estimation results from the model was then tested against the actual investigation data in 2014 Ludian earthquake-hit area. Results showed that, the trapped personnel distribution assessed by this model is generally concordant with that obtained by the actual survey in Ludian earthquake. The grid-based assessment of people trapped in earthquakes, can meet the requirements of key search and rescue zone identification and rescue forces allocation in the early stage of earthquake emergency. Our results also showed that there are many factors affecting the distribution of people trapped in earthquakes. The population distribution and the collapse of buildings are the core factors. Additionally, the factors related to human response (e.g., personal protection actions, etc.) also have important effects on the number of people trapped in earthquakes. LBS technology can almost real-time monitor the change of regional population, and further enhance the evaluation accuracy of people trapped in an earthquake. Although there were some limitations in the study, it offers a rapid approach for assessing the trapped people losses. The approach can be further improved to provide more information and suggestions for earthquake emergency search and rescue.

Space-borne sensors are producing a large number of remotely sensed data, and, consequently, various measurements of the same field are available to end-users. Alternatively, various satellite processing centers are producing extensive products based on data of only one mission. This is precisely the case of the Gravity Recovery and Climate Experiment (GRACE) mission, which has been monitoring terrestrial water storage (TWS) since April 2002, while the Center for Space Research (CSR), the Jet Propulsion Laboratory (JPL), and the Goddard Space Flight Center (GSFC), provide individual monthly solutions in the form of mass concentration (mascon) “blocks”. The inverted surface densities maps are being used in many applications, and therefore, as no ground truth data exist, the uncertainties are unknown. The assessment of the uncertainties associated with these various products is mandatory to guide data producers and support the users to choose the best dataset. However, uncertainties estimation of space-borne products often relies on ground truth data, and, in the absence of such data, the assessment of their qualities is a challenge. Here, it is proposed to estimate the quality of mascon solutions from CSR, JPL, and GSFC to guide users and data producers on the overall assessment of the products.

The temperature in subsurface porous media (subsurface temperature) has been popularly considered as a natural tracer to assess the groundwater flux in aquifers. Previous studies usually consider steady-state heat transfer and ignore the boundary effects on the subsurface temperature to simplify the mathematical models. However, the simplified models are expectedly inapplicable to the cases of transient temperature in porous media induced by boundary effects. This study proposes an analytical heat- transfer model for portraying the transient subsurface temperature profiles due to convective boundary effects. The model is composed by a heat transfer equation subject to a convective boundary condition. The surface temperature and initial subsurface temperature are both assumed constant. The study results show that the convective boundary effects on the subsurface temperature are significantly affected by the thicknesses of the media, thermal property of the convective boundary, and the groundwater flux.

The Lanyang Plain is located at the northeastern part of Taiwan, which is adjacent to the Pacific Ocean and the Hsuehshan Range, Central Range is in its south to west. The study area is located in Da-yin area which is between Lanyang River and Luodong River, which brings thick gravel layers. This area is a great potential for groundwater developments, so we tried to use geophysics method to know the hydrogeological parameters of this area. In the study, We applied 2D Electrical Resistivity Tomography (ERT) method to estimate the depth of groundwater table and specific yield of the unconfined aquifer in dry and wet seasons. We have deployed ten survey lines in the study area. And the investigation is conducted near the Da-yin groundwater observation well to compare with groundwater estimation. Based on Archie's Law, we converted the resistivity data into relative water contents. Using the Van-Genuchten model to characterize the depth of the water table, and calculate the specific yield by the difference between the saturated and residual water contents. Finally, the results can depict the water level distribution in the study area. According to the results, the specific yield is 0.08 to 0.15 and the groundwater level decreases from southwest to northeast, which is same as the flow direction of Luodong River. It is speculated that this area is affected by the rainfall from mountain and river recharge.

Land surface temperature (LST) is a key climate observable used detect changes in the Earth’s surface energy budget. Land surface temperature exhibits strong diurnal cycling and spatial heterogeneity. The choice of orbital pattern in satellite remote sensing entails a tradeoff between high spatial resolution (low earth orbit, LEO) and high revisit frequency (geostationary, GEO), in which no single sensor combines the best achievable performance in both domains. As a result, data fusion techniques have long been used to produce synthetic data products combing the strengths of multiple datasets. Here, we exploit co-located, co-temporal observations from GEO and LEO satellites to develop a deep learning-based method for GEO-LEO super-resolution. We use GeoNEX-L1G multispectral observations from GOES-16 alongside a 70m LST product from the ECOTRESS sensor. Applying a dual network approach, we predict LST at the native GOES-16 resolution with <2 K error, then downscale LST using low- to high-resolution mappings learned from ECOSTRESS. Comparison is presented with other satellite-based land surface temperature datasets (NOAA GOES-R LST, MOD11, MYD11) and ground observations from the Surface Radiation (SURFRAD) Network. We anticipate that the synergies between a variety of active orbit configurations can be used to overcome traditional constraints on retrieval capabilities and establish a framework for enhanced monitoring of spatial and temporal variability across multiple geophysical variables.

GeoNEX is a collaborative project led by scientists from NASA and many other international institutes to generate Earth monitoring products using data streams from the latest geostationary (GEO) sensors. Its consistent processing and common gridding systems can produce research-quality data products from GEO sensors and leverage GEO-GEO or GEO-LEO (low earth orbit) synergistic uses. Currently, GeoNEX has produced and disseminated L1G (geometrically corrected Level 1 products) from GOES 16/17 ABIs and Himawari-8 AHI, but a new Korean geostationary sensor (Advanced Meteorological Imager, AMI) onboard Geo-KOMPSAT-2A covering a large proportion of Asia and all of Oceania is in development. Our recent efforts on assessing geolocation accuracy in ABI and AHI suggest a nontrivial residual exists in both level 1B data with varying spatiotemporal patterns. The findings urge us to prioritize identifying and correcting geolocation residuals of AMI to generate accurate and consistent GeoNEX top-of-atmosphere (TOA) reflectance products and following processing chains. Here we implement a phase correlation correction approach to a visible band (0.64 µm, 500 m) using landmarks prepared from finer scale digital terrain models. We characterize spatiotemporal patterns (e.g., diurnal & daily) of geolocation residuals of AMI before and after correction. The geolocation corrected AMI data are further compared with GeoNEX AHI L1G products which are able to create unique stereo-type observations with AMI through leveraging the similarities of spectral bands and the sun-target-sensor geometry. Further, we discuss challenges in utilizing the GEO-GEO (e.g., AMI & AHI) satellite data for potential applications.

Globally, there are now hundreds of ground-based environmental monitoring stations routinely collecting data on a variety of earth-atmosphere interactions. Such observations are also being augmented with data from orbiting satellites. With the beginning of the EOS-era, the MODIS subset around flux towers has been frequently used for validating ecosystem models developed at flux towers and upscaling the observed flux data to regional scales. However, MODIS on the polar-orbiting satellites can observe target regions only once a day, while the Fluxnet eddy-covariance data are compiled as sub-hourly. Therefore, summarizing the sub-hourly flux data into daily statistics is necessary for the comparison between MODIS and Flux data. The new generation geostationary satellite sensors (GOES-16/17 ABI and Himawari-8/9 AHI) have capabilities similar to MODIS but collect data at 5-15 minute intervals. These high-frequency observations allow us to understand and scale diurnal fluxes. Some studies have already shown the effective utilization of time series of geostationary satellite data for ecosystem modeling. We are producing NEX Level-1G products, which are gridded Top-of-Atmosphere reflectance and brightness temperature data from geostationary satellite sensors. We cut out the NEX Level-1G data using the same file format with the MODIS subset except for the projection. The other data products (e.g., surface reflectance, land surface temperature, vegetation indices, and climate data) will be added upon their availability. Currently included networks are Fluxnet, PhenoCam, and AERONET. The NEX subset data will be provided through NASA NEX data portal.

Observations from the third-generation geostationary satellite instruments (GOES 16/17 ABI, Himawari 8/9 AHI, and etc.) have spatial resolution and spectral band configurations comparable to flagship LEO sensors (e.g., MODIS/VIIRS). More importantly, these data are acquired at very high temporal resolution, faithfully recording the variations of the full disk of Earth at every 5-10 minutes. They thus provide unique information about Earth’s atmosphere and surface, which can be exploited in atmospheric correction algorithms. This study systematically analyzes the diurnal variability in the GeoNEX L1G TOA reflectance products and compares them to simulated results by state-of-the-art radiative transfer codes. Our results show that The smoothness of the TOA reflectance diurnal cycle provides a convenient and reliable way to identify stable atmospheric conditions and filter out passing clouds/shadows. The diurnal variability of the blue band (0.47µm) reflectance is regulated mainly by atmospheric optical conditions over a majority of land cover types. As such, the diurnal variability of the blue band data allows us to retrieve AOD without invoking the use of spectral band ratios (SRC) as in previous algorithms. In comparison, the diurnal variability of the short-wave infrared band (2.2µm) BRFs is mainly regulated by surface reflectance and the sun-target-satellite geometry. This information allows us to test and, if suitable, retrieve surface BRDF parameters. Spectral band ratios, especially those between the 2.2µm and 0.47µm bands, appear to converge to the range from 0.3 to 0.5 over dense-dark vegetation pixels but otherwise vary substationally depending on locations and sun-target-satellite geometries. As such, our analysis clearly demonstrates that the information content of the high-frequent diurnal geostationary observations is unique and thus helps us improve the performance of atmospheric correction algorithms.

Satellite-based fire observations have been widely used to investigate global biomass burning (BB) emissions during the past few decades. One of the observations is fire radiative power (FRP) that measures the instantaneous fire radiative energy emitted from actively burning fires. Because FRP is directly related to the rate of biomass consumption and smoke emissions, FRP observations from polar-orbiting and geostationary satellites have been increasingly applied to estimate BB emissions. Nevertheless, a polar-orbing satellite (e.g., MODIS, VIIRS) can only observe fire activity twice a day, while legacy geostationary satellites (e.g., MTSAT, GOES-11 and GOES-15) usually miss small and cool fires due to their coarse spatial resolutions. As a result, FRP observations from these satellites largely limits the accurate estimates of BB emissions. The new-generation geostationary satellites observe fires at unprecedented spatiotemporal resolutions, which highly improve the quantification of FRP dynamics. This study investigates FRP diurnal cycle using the 2-km GOES-16 and GOES-17 ABI active fire products that detect fires across the Conterminous United States (CONUS) every five minutes. Specifically, ABI fire detections from the two satellites are first merged based on view angle and further stratified by five main land cover types in different ecoregions. Then, the associated FRP observations are aggregated in 3km grids for each ecosystem in a given ecoregion. Finally, FRP diurnal cycles are derived using the aggregated FRP. Our preliminary results show that FRP diurnal cycle varies largely among different land cover types and differs significantly for the same cover type across ecoregions. The derived FRP diurnal cycles are expected to greatly enhance the estimation of hourly BB emissions and increase the accuracy of air quality forecast.