Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Copy DOI

For grid stability, operation, and planning, solar irradiance forecasting is crucial. In this paper, we provide a method for predicting the Global Horizontal Irradiance (GHI) mean values one hour in advance. Sky images are utilized for training the various forecasting models along with measured meteorological data in order to account for the short-term variability of solar irradiance, which is mostly caused by the presence of clouds in the sky. Additionally, deep learning models like the multilayer perceptron (MLP), convolutional neural networks (CNN), long short-term memory (LSTM), or their hybridized forms are widely used for deterministic solar irradiance forecasting. The implementation of probabilistic solar irradiance forecasting, which is gaining prominence in grid management since it offers information on the likelihood of different outcomes, is another task we carry out using quantile regression. The novelty of this paper lies in the combination of a hybrid deep learning model (CNN-LSTM) with quantile regression for the computation of prediction intervals at different confidence levels. The training of the different machine learning algorithms is performed over a year’s worth of sky images and meteorological data from the years 2019 to 2020. The data were measured at the University of French Polynesia (17.5770° S, 149.6092° W), on the island of Tahiti, which has a tropical climate. Overall, the hybrid model (CNN-LSTM) is the best performing and most accurate in terms of deterministic and probabilistic metrics. In addition, it was found that the CNN, LSTM, and ANN show good results against persistence.

This paper describes an original model based on The Energetic Macroscopic Representation (EMR) to develop a control structure of an innovative compression-assisted hybrid thermochemical cooler driven by low-grade thermal energy. Hybrid thermochemical cooler with mechanical compression have a wider operating temperature range than conventional thermochemical processes. This makes it possible to exploit lower grade heat sources and/or produce cold at lower temperatures. In addition, the ability to control compressor speed increases thermochemical process controllability to respond to load or source variations. Nodal modeling of each process component is developed according to the EMR formalism. A parametric identification and validation of this process model is then carried out using experimental data. The deviation from the experimental data is lower than 1 bar for the reactor, condenser and evaporator pressure, lower than 0.1 for the reaction advancement and lower than 2 degrees C for the reactor wall temperature. The EMR model is then inverted to obtain a process control law for maintaining a cold room at-18 degrees C. Despite the variation in cooling demand from a minimum of 100 W to a maximum of 700 W, the compressor's control structure was able to regulate its speed of rotation to maintain the cold room temperature at-18 +/- 0.5 degrees C for 24 h.

Many developments in recent years concern the hybridization of sorption systems with a mechanical compressor. This paper presents an assessment study of different hybrid thermochemical systems used to ensure a cold production. A screening over more than 100 ammonia salts have been done to investigate the different compressor configurations. The compressor can be used to assist the decomposition phase, the synthesis phase or both phases. Two major points are tackled: the modification of the operating range and of the performances of such hybrid systems. This article is intended to be a thermodynamic tool for selecting the most suitable ammoniated salt and compressor operating configuration for a targeted cooling application. Compression-assisted decomposition has already been studied in the scientific literature for thermochemical systems, but not the compression-assisted synthesis for a low-temperature cold production. This latter hybridization extends the operating range of the thermochemical process quite the same way as the assisted decomposition but brings better average performances over the operating range compared to a fully thermally driven system, with higher COP (+8%), exergetic efficiency (+2%) and primary energy efficiency (+4%). The possibility to use a compressor in both phases leads to very interesting performances over a largely extended operating range. In this case, the average exergetic efficiency can be 32% higher than the average exergetic efficiency of a fully thermally driven system.

In order to decarbonize electricity production in insular tropical regions, hydrogen as an energy vector appears to be a promising solution. But some issues have to be dealt with, like the overall yield of the hydrogen chain. Moreover, thermo-chemical systems can produce cooling by thermal recovery. In this paper, we study a system composed by an electrolyzer, a hydrogen fuel cell, a Li-ion battery pack and a thermochemical reactor coupled with a conventional ammonia heat pump. In this system, the waste-heat from the electrolyzer and the fuel cell are used to desorb a thermochemical reactor for a differed production of cooling. A Mixed Integer Linear Program is used to optimize the energy management, taking into account the electrical and thermal load demand and the aging of the electrolyzer, the fuel cell and the battery. Results show that the system is able to provide the electrical and thermal needs of the load and the use of the thermochemical cooling system improves the fuel cell and the electrolyzer efficiency by 11 % and 15 % respectively compared to a system without thermochemical storage.

Solar-power-generation forecasting tools are essential for microgrid stability, operation, and planning. The prediction of solar irradiance (SI) usually relies on the time series of SI and other meteorological data. In this study, the considered microgrid was a combined cold- and power-generation system, located in Tahiti. Point forecasts were obtained using a particle swarm optimization (PSO) algorithm combined with three stand-alone models: XGboost (PSO-XGboost), the long short-term memory neural network (PSO-LSTM), and the gradient boosting regression algorithm (PSO-GBRT). The implemented daily SI forecasts relied on an hourly time-step. The input data were composed of outputs from the numerical forecasting model AROME (Météo France) combined with historical meteorological data. Our three hybrid models were compared with other stand-alone models, namely, artificial neural network (ANN), convolutional neural network (CNN), random forest (RF), LSTM, GBRT, and XGboost. The probabilistic forecasts were obtained by mapping the quantiles of the hourly residuals, which enabled the computation of 38%, 68%, 95%, and 99% prediction intervals (PIs). The experimental results showed that PSO-LSTM had the best accuracy for day-ahead solar irradiance forecasting compared with the other benchmark models, through overall deterministic and probabilistic metrics.

There was an error in the original publication [...]

The South Pacific Tropical Upper Tropospheric Trough (TUTT) is an elongated trough that appears in summer monthly averaged maps of the upper tropospheric flow over the ocean. We provide the first comprehensive description of the South Pacific TUTT and analyze its impact on the regional climate with 42 yr (1979-2020) of monthly data from ERA5, radiosonde, rain, keraunic data, and lightning flash rate from the Optical Transient Detector-Lightning Imaging Sensor. The data confirm the presence of the TUTT along a southeast-northwest axis from November to May. Located between 100 and 400 hPa, the TUTT is a cold-core with a relative vorticity minimum and a negative geopotential anomaly. Divergent, ascending flow with high relative humidity is found south and west of the TUTT axis, while convergent, descending flow with low relative humidity is observed north and east of the TUTT axis. The seasonal and long-term impacts of the TUTT on the local and regional climate is dependent on its location and its strength. The existence of the TUTT in the South Pacific is strongly dependent on the El Niño-Southern Oscillation (ENSO) phases, being stronger during a strong negative ENSO phase and disappearing from French Polynesia during a strong positive ENSO phase. Genesis of tropical disturbances east of 180°W is possible only if the TUTT is located east of 120°W. The environmental conditions associated with the TUTT, along with lightning and rainfall data from 3 sites in French Polynesia, show that lightning flash rates are higher during a negative ENSO phase than during a positive ENSO phase.

Tropical climate is characterized by hot temperatures throughout the year. In areas subject to this climate, air conditioning represents an important share of total energy consumption. In some tropical islands, there is no electric grid; in these cases, electricity is often provided by diesel generators. In this study, in order to decarbonize electricity and cooling production and to improve autonomy in a standalone application, a microgrid producing combined cooling and electrical power was proposed. The presented system was composed of photovoltaic panels, a battery, an electrolyzer, a hydrogen tank, a fuel cell, power converters, a heat pump, electrical loads, and an adsorption cooling system. Electricity production and storage were provided by photovoltaic panels and a hydrogen storage system, respectively, while cooling production and storage were achieved using a heat pump and an adsorption cooling system, respectively. The standalone application presented was a single house located in Tahiti, French Polynesia. In this paper, the system as a whole is presented. Then, the interaction between each element is described, and a model of the system is presented. Thirdly, the energy and power management required in order to meet electrical and thermal needs are presented. Then, the results of the control strategy are presented. The results showed that the adsorption cooling system provided 53% of the cooling demand. The use of the adsorption cooling system reduced the needed photovoltaic panel area, the use of the electrolyzer, and the use of the fuel cell by more than 60%, and reduced energy losses by 7% (compared to a classic heat pump) for air conditioning.

In order to establish data to be used for comparison with future evolution related to climate change, physical, chemical and biological characteristics of dew and rainwater as collected during the 2005 dry season (plus a few data during the 2004 dry season) are reported. They have been collected in two characteristic tropical islands of French Polynesia, Tikehau (TKH), a low-lying coral atoll in the Tuamotu Archipelago, and the mountainous Tahiti island at the University of French Polynesia (TAH). Trade winds dominate the trajectory of air masses, ensuring constant temperature and humidity to the lower layers of the atmosphere where dew forms. In addition to the comparison of dew yields with a physical model using simple meteorological data (air and dew point temperatures, windspeed, cloud cover), the following parameters were studied: pH, electrical conductivity (EC), total dissolved solids (TDS), total hardness, suspended matter, ion concentration with major cations (Ca2+, K+, Na+, Mg2+, NH4+) and major anions (Cl−, SO42−, NO3−), reviviscible aerobe microorganisms and compared to two Polynesian spring waters (“Eau Royale” and “Vaimato”). Dew, with a chemical composition mainly consisting of Na+, Ca2+, Mg2+, Cl− and SO42−, exhibits much higher ion concentration than rain and compares well with the composition of local spring waters. The values of pH, EC and TDS are larger in dew than in rainwater. At TAH, the volume weighted mean (VWM) pH values of dew were 6.05 in 2004 and 5.23 in 2005, larger than the pH of rain (4.69). The VWM dew EC (at TAH, 203 μS·cm−1 in 2004 and 237 μS·cm−1 in 2005; 321 μS·cm−1 at TKH in 2005) and dew TDS (152 mg·L−1 at TAH and 225 mg·L−1 at TKH) were higher than the corresponding quantities in rain. Mean total hardness (TH) values of dew water are much higher than in rainwater (2.8 for dew versus 0.5 for rain at TAH and 5.7 for dew versus 0.5 for rain at TKH). Ions Na+, Mg2+ and Cl− are clearly of sea origin while the presence of Ca2+ is due to coral particles. From their chemical characteristics, dew and rainwater could be used as an alternative source of water in dry season but, due to the presence of reviviscible aerobe microorganisms at 22 °C and 36 °C (>300 CFU·mL−1), water must be disinfected to be potable.

Three different studies are presented in this paper. As a first step, a Particle Swarm Optimization (PSO) algorithm is used to optimize a prototype of cold/electricity cogeneration designed to be disconnected from the grid and implanted in an insular tropical region where a high need of cold and electricity is required. The electricity is provided by solar photovoltaic panels and the electrical energy in excess is stored in the form of hydrogen thanks to an electrolyzer. When a lack of electricity occurs a fuel cell provides the missing electricity by using the stored hydrogen. An electrically driven heat pump is also used to produce and cover the cold needs. Finally, in order to increase the overall efficiency of this electricity/cold cogeneration system, the low-grade waste heat generated by the different components of the system, mostly the electrolyzer and the fuel cell, is recovered and upgraded by a thermochemical reactor enabling a further cold production. The thermochemical reactor assists the heat pump for the cold supply, decreasing thus the electricity consumption. Such a prototype is intended to be built in Tahiti during the RECIF project. The PSO algorithm has been implemented and results are promising because component’s size is reasonable, and the driving strategy is consistent while both demands are always satisfied. In a second study, the same PSO algorithm has been used to perform an analysis and identify a general shape of load profiles for which it become interesting, from an economic point of view, to store electricity into hydrogen instead of electrochemical batteries when the cold production is only handled by a heat pump. This study has shown that the more electricity is consumed at night, the more it is interesting to use hydrogen. Finally, the algorithm has been used to see the evolution of the economic interest when a thermochemical system is added. This study has been carried out considering a storage into hydrogen and an exploitation of the low-grade heat by a thermochemical unit to enable a further cold production. The economic interest of a thermochemical system has not been proven by this study, that is why further considerations must be considered in order to justify its use as ecological impacts for example.

This study focuses on the solar resource available at Faaa, Tahiti (17.5°S, 149.5°W) thanks to 10 year-long solar irradiance time series. Faaa’s global horizontal irradiance ranges from 14 MJ.m-2.day-1 (June) to 21 MJ.m-2.day-1 (November) in agreement with the sun’s annual path, while clearness index ranges from 0.5 (January) to 0.67 (July), in agreement with the wet and dry seasons. The Global Solar Atlas satellite-derived dataset shows acceptable relative error when compared to Faaa in situ measurements. This product could then be used for other coastal areas of Tahiti. The annual energy output of a single PV module is 256.7 kWh, which corresponds to 7 % of the annual consumption of a typical household in Tahiti. The capacity factor reaches 22.5 %, which makes Faaa a good site for harnessing solar resource.

Intraseasonal and diurnal variability of precipitation over Tahiti, French Polynesia, are investigated with the use of wind regimes as intermediary tools. Four wind regimes have been determined over a large domain around Tahiti in the wet season. It has been shown that some phases of the Madden–Julian Oscillation (MJO) trigger some regimes and undermine some others. The diurnal cycle of precipitation in Tahiti is composed of two rainfall maxima. A frequent maximum at midnight but with a low value. An afternoon maximum, less frequent but with a higher value. We have noticed that the afternoon rainfall maximum, which is a reasonable consequence of the sea breeze circulation, is smoothen on the windward side when vigorous wind regimes obstruct the development of the sea breeze. We have also found that daily mean precipitation is significantly increased during MJO Phases 7 and 8, and reduced during Phases 2 and 3. Phase 8 is the wettest phase on Tahiti, suggesting the convective envelope is overlooking the island. Lastly, we have provided evidence that the diurnal cycle of rainfall is significantly affected by the MJO. Indeed, both the midnight and afternoon maxima increase during MJO Phase 7. The most quiescent wind regimes are triggered during Phase 7, as a result from the propagation of the westerly wind anomaly, which allow the development of the sea breeze circulation and the increase of the afternoon peak. For MJO Phase 8, however, the above relationships break down, as the diurnal cycle is not enhanced by the predominant quiescent wind regimes. We assume that the MJO convective envelope covers the island and stays several days, preventing the solar heating of the surface, hence the development of the land/sea breeze circulation, and foremost leads to higher amounts of rainfall as the Phase 8 is the wettest phase on Tahiti.

The energy situation in tropical insular regions, such as in the French Polynesian islands, presents a number of challenges, including high dependence on imported fuel, high transport costs from the mainland and weak electricity grids. With regards to electrical energy demand, the high temperatures in these regions throughout the entire year implies that a large proportion of electricity consumption (~40%) is used to cool buildings, even during evening hours. This paper presents an air conditioning system driven by photovoltaic (PV) electricity that combines a mechanical vapor refrigeration system and a thermochemical storage unit. Thermochemical processes are able to store energy in the form of chemical potential with virtually no losses, and this energy can be used to produce cooling during the evening hours without the need to run a compressor. The efficiency of such a hybrid system is evaluated and compared with alternative processes that utilize either electrochemical (Pb, Li-ion batteries) or thermal storage (ice, chilled water) for cooling production.

Performance assessment of buildings in tropical climates requires annual meteorological data files :for an "energy approach" of air-conditioned buildings. It also requires localized climatic sequences allowing a real estimation of the natural ventilation potential of the sites for a "comfort" approach of non-air-conditioned buildings. The absence of this type of weather data in Polynesia considerably limits designer analysis means. This article presents two methods to establish meteorological sequences for both energy and comfort approaches. The first aims developing annual typical weather files from a reduced number of ground measurements. The missing meteorological variables are completed using a global to diffuse decomposition model of the solar irradiance. The second is based on a mesoscale climate model and a downscaling to obtain a characterization of localized natural ventilation potential and climatic data on a fine mesh. The weather sequences generated are used to perform energy calculation for air conditioned classrooms and wind potential and comfort assessment for cross ventilated classrooms of a primary school located in French Polynesia.

The Earth's naturally occurring Schumann resonances (SR) are composed of a quasi-continuous background component and a larger-amplitude, short-duration transient component, otherwise called 'Q-burst' (Ogawa et al., 1967). Sprites in the mesosphere are also known to accompany the energetic positive ground flashes that launch the Q-bursts (Boccippio et al., 1995). Spectra of the background Schumann Resonances (SR) require a natural stabilization period of similar to 10-12 min for the three conspicuous modal parameters to be derived from Lorentzian fitting. Before the spectra are computed and the fitting process is initiated, the raw time series data need to be properly filtered for local cultural noise, narrow band interference as well as for large transients in the form of global Q-bursts. Mushtak and Williams (2009) describe an effective technique called Isolated Lorentzian (I-LOR), in which, the contributions from local cultural and various other noises are minimized to a great extent. An automated technique based on median filtering of time series data has been developed. These special lightning flashes are known to have greater contribution in the ELF range (below 1 kHz) compared to general negative CG strikes (Huang et al., 1999; Cummer et al., 2006). The global distributions of these Q-bursts have been studied by. Huang et al. (1999) Rhode Island, USA by wave impedance methods from single station ELF measurements at Rhode Island, USA and from Japan Hobara et al. (2006). The present work aims to demonstrate the effect of Q-bursts on SR background spectra using GPS time-stamped observation of TLEs. It is observed that the Q-bursts selected for the present work do alias the background spectra over a 5-s period, though the amplitudes of these Q-bursts are far below the background threshold of 16 Core Standard Deviation (CSD) so that they do not strongly alias the background spectra of 10-12 min duration. The examination of one exceptional Q-burst shows that appreciable spectral aliasing can occur even when 12-min spectral integrations are considered. The statistical result shows that for a 12-min spectrum, events above 16 CSD are capable of producing significant frequency aliasing of the modal frequencies, although the intensity aliasing might have a negligible effect unless the events are exceptionally large (similar to 200 CSD). The spectral CSD methodology may be used to extract the time of arrival of the Q-burst transients. This methodology may be combined with a hyperbolic ranging, thus becoming an effective tool to detect TLEs globally with a modest number of networked observational stations.

Assessing natural ventilation potential on a tropical island such as Tahiti (French Polynesia), where strong climatic constraints apply on buildings constitute a challenge due to the scarcity of wind measurements. Indeed, only one in situ automatic weather station provides high quality wind speed and direction data records. To overcome this lack of information, a dynamical downscaling using WRF-ARW model has been performed to assess 10 m wind speed and direction at high spatial resolution in Tahiti. A weather type classification is used to highlight the main regimes prior to the downscaling process. First of all, daily 700 hPa geopotential and 10 m horizontal wind components from reanalysis dataset undergo a clustering technique resulting into six wind classes. Then, the five closest dates to the center of each wind class are selected. Simulations with the model WRF-ARW are then performed for selected days over 3 interactively nested domains over French Polynesia, with finest horizontal mesh size of 1.33 km over Tahiti Island. The initial and coupling fields are derived from the ERA Interim reanalysis dataset. The only one in situ station is then used to assess the performance of the downscaling. The higher resolution obtained with this model setup allowed to highlight the contrast between the leeward and windward side of the island. Better resolving the complex topography of the volcanic island, these high resolution recurrent wind regimes would be useful to support low environmental impact construction policies.

The PEACH project (Projet en Electricité Atmosphérique pour la Campagne HyMeX – the Atmospheric Electricity Project of the HyMeX Program) is the atmospheric electricity component of the Hydrology cycle in the Mediterranean Experiment (HyMeX) experiment and is dedicated to the observation of both lightning activity and electrical state of continental and maritime thunderstorms in the area of the Mediterranean Sea. During the HyMeX SOP1 (Special Observation Period) from 5 September to 6 November 2012, four European operational lightning locating systems (ATDnet, EUCLID, LINET, ZEUS) and the HyMeX lightning mapping array network (HyLMA) were used to locate and characterize the lightning activity over the northwestern Mediterranean at flash, storm and regional scales. Additional research instruments like slow antennas, video cameras, microbarometer and microphone arrays were also operated. All these observations in conjunction with operational/research ground-based and airborne radars, rain gauges and in situ microphysical records are aimed at characterizing and understanding electrically active and highly precipitating events over southeastern France that often lead to severe flash floods. Simulations performed with cloud resolving models like Meso-NH and Weather Research and Forecasting are used to interpret the results and to investigate further the links between dynamics, microphysics, electrification and lightning occurrence. Herein we present an overview of the PEACH project and its different instruments. Examples are discussed to illustrate the comprehensive and unique lightning data set, from radio frequency to acoustics, collected during the SOP1 for lightning phenomenology understanding, instrumentation validation, storm characterization and modeling.