Higher GHI than GTI is a normal scenario. This happens when direct irradiance is so low such that the benefits of being tilted towards the sun are offset by the reduction in visible sky which decreases the amount of incident diffuse radiation. This phenomena is why smart tracking panels may rotate to a horizontal position under cloudy conditions to maximise the amount of sky they can see.

GTI can be referred to as the tilted irradiance.

GHI or DHI being larger than DNI is not an uncommon scenario. This happens under partially cloudy conditions when there can be little or no direct irradiance (so DNI is low or 0) but there is still some diffuse irradiance which causes GHI to be larger.

Cloud_opacity10, ghi10, dni10 and dhi10 form a physically consistent "possible cloudier sky" scenario based on the 10th percentile clearsky index (where clearsky index = 1 - cloud_opacity / 100). This means that physically meaningful PV power modelling can be carried out using ghi10, dni10 and dhi10 parameters. The same is true of cloud_opacity90, ghi90, dni90, and dhi90 for the "possible clearer sky" scenario based on the 90th percentile clearsky index.

In practice, this means that the total global irradiance follows the intuitive relationship ghi10 <= ghi90, since the "90th percentile clearness" scenario has less cloud than the "10 percentile clearness".

However, at low to moderate cloud opacity, sometimes dhi10 >= dhi90. For example, the DHI (diffuse irradiance) can actually be higher than the clearsky scenario, as sunlight in the direct beam (dni) is scattered throughout the cloud and to the ground. This is why glare can actually be worse when it is overcast - if the cloud is not too thick, a section of cloudy sky can actually be brighter than the same section of blue sky, even though the total irradiance reaching the ground is smaller.

It is the average value of the time period. All our parameters are averages over the period.

Solcast continues to update our estimates as new data sources become available. This includes historical estimates in our live and historic endpoints. The most impactful of these new data sources is newer satellite imagery becoming available. This typically occurs within 30 minutes of the current time but this may extend to a few hours in situations such as interruptions to data feeds.

If you are building a historical record from live data record, we would suggest making one API call for the previous day, as by then all satellite imagery, the primary driver of irradiance variability, will have safely been assimilated. If you need data earlier but want to ensure fewest number of changes we suggest making the request as late as possible.

All Solcast irradiance estimates are broadband.

The Solcast albedo value provided is a daily value and does not represent any diurnal angular dependence of reflectivity. This value is derived from a dataset with a resolution of around 50km. Thus depending on the size of the location, the albedo value is likely to be the same throughout. Apart from diurnal variations due to angular dependence, albedo will tend to vary gradually. For example, as vegetation state changes with the seasons. Note that albedo is difficult to measure in detail, especially with variation in the exact nature of ground cover.

To accurately convert horizontal and direct irradiance to tilted irradiance, Solcast recommends using a transposition model.  This takes account of factors such as the reflected irradiance striking the panel from the ground. Information regarding the transposition models we use at Solcast for calculating tilted irradiance, with references, is available at Irradiance Data Methodology | Solcast™

This is due to the inverters being undersized compared to the solar panels, and thus the DC/AC ratio being greater than 1. The maximum power is clipped by the AC capacity.

The albedo parameter available from the Solcast API are daily estimates of albedo and derived from MERRA2 based estimates.

Solcast also uses albedo within our clear sky model, further information is available here: https://solcast.com/irradiance-data-methodology and in its advanced PV power model as a site specific configuration.

Alternatively; Why doesn't my long-term historical GHI match the TMY GHI average?

The TMY is constructed so that the weighted average:

AVERAGE(ghi_weight*GHI + dni_weight*DNI)

matches between the TMY and the long-term historical timeseries.

Set ghi_weight=1, dni_weight=0 to make the TMY GHI average match the long-term historical timeseries average.

You may want to use the default (ghi_weight=0.2, dni_weight=0.8) because for some applications, such as tracking solar systems, DNI (direct normal irradiance) can be more predictive of power production.

Do the time periods used match? The TMY calculation uses historical timeseries up to the end of the last calendar year, updated yearly. This ensures equal samples of all calendar months go into the TMY calculation and that we have the best possible input data. If you compare the TMY to a different time-period, the average irradiance values may be different.