Site-adaptation of Solargis data

Solar resource historical database provides a unique resource for elaborated climate statistics that can help understand solar energy resource for any particular site. In most places, where there is no availability of ground stations next to the project site, Solargis satellite-based model provides a consistent, cost effective and gap-free multi-year data period that can be used for solar resource assessment.

On the other hand, weather parameters retrieved from satellite-based meteorological models have lower spatial and temporal resolution compared to on-site meteorological measurements. Therefore modelled parameters may characterize certain climate patterns at the surrounding area level rather than showing very local microclimate conditions.

As a standard practice, a meteorological station is deployed at a site of large solar energy project development. Deployment of solar measuring stations in a country has the strategic advantage of adapting and validating the radiation model at regional level to provide high-quality data and information for decision-makers and investors.

Running a ground-monitoring campaign and combining measurements with satellite data is the way to maintain low uncertainty of solar resource at a project site in the long term:

• During the planning phase, the main objective of measuring data at the project site is to record accurate local meteorological characteristics and use them in the adaptation of the satellite-based model to reduce uncertainty of the long-term time series and aggregated estimates.
• During the plant’s operation phase, on-site measuring is relevant for accurately measuring plant’s performance and detection of failures. Here satellite time series should be used as an independent source of information for quality control and optimization of ground-measured data.

The data correlation is effective for mitigating systematic problems in the satellite-derived data (e.g. under/over-estimation of local aerosol loads) especially when the magnitude of the deviation is invariant over the time or has a seasonal periodicity. The accuracy-enhancement methods are capable to adapt satellite-derived DNI and GHI datasets (and derived parameters) to the local climate conditions that cannot be recorded in the original satellite and atmospheric inputs.

Satellite-based Solargis data can be adapted to the project site when at least 12 months of ground-measurements are available. The result of this process is the construction of a multi-year solar dataset with improved accuracy.

For the adaptation of satellite data to the conditions represented by the ground measurements at the project site, two main approaches are taken:

• Adaptation of satellite-based GHI and DNI values. Using this method the bias (systematic deviation) is corrected together with fitting the cumulative distribution functions.
• Adaptation of the input parameters and data used in the solar radiation model. More complex parameters, such as Aerosol Optical Depth and/or Cloud Index are adjusted using this approach.

As developers of the full computational chain, in Solargis we have the capacity of adapting the model input data, so both methods can be combined for achieving consistent and accurate results. Other methods only using a statistical approach will achieve not so good results on accuracy.

The adaptation of Solargis input parameters are used for correcting the main sources of discrepancies (such as limitations in aerosol description). Small residual deviations are removed in the next step by a simpler adaptation of the output values. Using this combined method for site-adaptation of satellite data, we are able to keep consistency of GHI, DNI and DIF components.

The data adaptation is important especially when specific situations such as extreme irradiance events are to be correctly represented in the enhanced dataset. These methods have to be used carefully, as inappropriate use for non-systematic deviations or use of less accurate ground data leads to accuracy degradation of the primary satellite-derived dataset.

A detailed data cleaning and quality control is always required as a first step when running site-adaptation. This process is based on SERI QC, BSRN and other in-house developed approaches. Time aggregation, harmonization and qualification are also required for next steps.

The ability to perform site-adaptation of satellite data is determined by several factors:

• Quality of sensors. It is recommended to use best category meteorological instruments for measuring GHI (secondary standard pyranometers and first class pyrheliometers). As a substitute to pyrheliometer RSR (Rotating Shadowband Radiometer) can be used, however uncertainty of measured GHI and DNI is higher. Use of redundant instruments (optimally one for each components: GHI, DIF and DNI) increases accuracy and reliability of the whole process.
• Quality of measurements. This is determined by regular maintenance, cleaning and calibration. A set of Quality Control routines, both automatic and managed by an operator, are to be evaluated. Only data which are pre-qualified may be used for site adaptation.
• Adequate length of ground measurements. Optimally, high-quality ground measurements should be available for a period of optimally 12 months. In case of a tight time schedule, a shorter period (9+ months) may be considered for the purpose of site-adaptation. However such data may not be capable to cover all seasonal deviations. Data covering a shorter period (e.g. 3-6 months) may provide a false indication of the relationship between long-term historical Solargis data and local measured information.