Satellite-based albedo data

Overview

The traditional assumption of surface albedo as a constant value in solar energy applications has been increasingly challenged with the advent of bifacial photovoltaic (PV) technology. This innovation relies heavily on accurate albedo data to maximize the harnessing of reflected solar energy. Albedo, a measure of surface reflectivity, directly influences the total reflected solar irradiance (RHI) and is a critical parameter for calculating tilted irradiance (GTI) and optimizing bifacial PV systems.

While ground-based measurements alone often prove inadequate, satellite-derived albedo estimates, such as those available through MODIS data products, offer a robust alternative. These satellite-based datasets have become integral to the solar industry, serving as a foundation for the development of precise albedo databases. Additionally, reanalysis datasets derived from numerical weather prediction (NWP) models provide gap-free albedo inputs, further enhancing data reliability.

By post-processing and combining these data sources, it is possible to derive reliable historical albedo values for specific locations. These values are typically provided in two formats: monthly averages or Time Series (daily values), depending on the analysis requirements. Monthly averages are generally recommended during the pre-feasibility stage, offering a broad overview, while Time Series data are more suited for the due diligence phase, providing detailed, temporal insights for advanced analysis.

Albedo Time Series calculation

Albedo is estimated based on surface reflectance data captured by instruments aboard satellite platforms. One such instrument is the Moderate Resolution Imaging Spectroradiometer (MODIS), which operates on the Terra and Aqua satellites. These satellites are strategically positioned in complementary orbits: Terra crosses the equator from north to south in the morning, while Aqua follows an opposite path, crossing from south to north in the afternoon. Together, they provide near-global coverage of the Earth's surface every 1 to 2 days.

Albedo is not typically provided directly as a standalone satellite product, even with MODIS data. Instead, it is derived from two distinct components:

To calculate the effective albedo value, the diffuse fraction (D2G) data from the Solargis irradiance database is applied, combining black-sky and white-sky albedo values. This integration allows for a more accurate representation of albedo under varying solar radiation conditions.

For Solargis Time Series of albedo, data sources that have a long historic record and reasonably high temporal resolution (at least daily values) are used. These include MODIS products MCD43A1 and MCD43A3. However, these products present missing data (data gaps), mainly due to persistent cloudiness.

Some examples of situations when the use of Time Series is more appropriate include:

• Satellite-derived albedo is to be compared with ground-based albedo measurements.

• Monthly average data may not be sufficient for considering individual stochastic events such as ephemeral snow. Only the averaged effect of such events is considered.

• The land cover of the project location may have changed over time, and an average of years 2006-2015 may not represent the current situation.

• The land cover at the project location is quite heterogeneous, and a dataset with a spatial resolution of 0.5 km would give more representative results than a dataset having 1 km resolution.

Gap-filling and monthly aggregation

Gaps in satellite-derived albedo data are unavoidable, often caused by persistent cloud cover or misdetections, and these gaps are reflected in historical time series datasets. Consequently, temporal and spatial gaps are common in time series data derived from MODIS. The size of these gaps is influenced by the dataset's spatial resolution, while their duration depends on the persistence of cloudy conditions.

Despite these challenges, monthly averages can be calculated to create gap-free datasets. This is achieved by combining pre-processed MODIS products with additional datasets from meteorological reanalysis databases, such as those from the ECMWF ERA-5 and NASA MERRA-2 projects.

The primary advantage of reanalysis-based albedo calculations is their ability to provide consistent, global data for any surface without gaps. However, the trade-off is the lower spatial resolution compared to other methods.

To calculate monthly averages of albedo, gap-filled snow-free MODIS product MCD43GF is used as the primary data source. However, since this MODIS product does not include albedo values for ephemeral snow, additional information from other MODIS products and ERA5 reanalysis data are combined to prepare a gap-free high-resolution dataset for all kind of land surfaces (including ephemeral snow). Finally, additional information from National Snow and Ice Data Center (NSIDC) is used to add the albedo of ocean ice (NSIDC). To correct imperfections and errors related to imprecisions of the MODIS land-sea mask we use additional layers from Solargis database.

The obtained dataset offers gap-free albedo information with high spatial resolution and global coverage, making it suitable for a variety of applications in solar energy and environmental modeling.

Final albedo products from Solargis have been validated using available ground measurements from albedometers. An albedometer comprises two pyranometers mounted in opposite directions, parallel to the horizontal surface, to simultaneously measure global horizontal irradiance (GHI) and reflected horizontal irradiance (RHI).

To enhance the accuracy of satellite-derived ground albedo data, dedicated ground measurement campaigns are often conducted for projects where albedo significantly impacts the expected energy yield. These campaigns provide localized, high-precision data, complementing satellite-based estimates.

However, due to the inherent temporal variability of albedo, measurement campaigns can yield widely dispersed results, influenced by the duration of the observation period. Short campaigns, lasting only a few days, often introduce greater errors than those found in satellite-based data, highlighting the importance of extended measurement durations for reliable validation.

Further reading

• Surface albedo and reflectance: Review of definitions, angular and spectral effects, and intercomparison of major data sources in support of advanced solar irradiance modeling over the Americas. C. Gueymard, V. Lara-Fanego, M. Sengupta, Y. Xie, 2019, Solar Energy, Volume 182, April 2019, Pages 194-212

• How to use albedo for more accurate bifacial PV estimates. White Paper. Solargis website, 2022.