Abstract
A common global vertical reference frame for physical heights is essential for many applications, such as studying and monitoring climate-related changes within the Earth system. The International Height Reference System (IHRS), defined by the International Association of Geodesy (IAG) in 2015, provides a global reference for physical heights. On a regional or local level, a combination of GNSS, a gravimetric (quasi)geoid model and precise levelling observations is an option to realise the IHRS. A previous study has shown that this so-called levelling-assisted realisation has the potential to significantly improve the standard (GNSS- and) model-based method in case a properly weighted least squares adjustment is made. This can result in a dense IHRF solution with reduced uncertainty. However, model-based geopotential numbers determined without levelling at a later stage will then not agree with the already existing levelling-assisted realisation. This disagreement is due to the smoothing effect obtained in the network adjustment process. The smoothing effect is caused by the low relative uncertainty of levelling observations, especially over short distances. In this contribution, we study this potentially systematic bias and how to construct a correction model to be applied to model-based geopotential numbers to improve their agreement with a weighted levelling-assisted IHRS realisation. This is investigated in a case study over Sweden. It is found that the proposed correction model does not significantly improve the systematic agreement between a model-based realisation made at a later stage with the levelling-assisted realisation in Sweden. However, it is concluded that it is important to be aware of that this type of correction might be needed in other regions. This should be investigated for any new levelling-assisted realisation.

This publication has 8 references indexed in Scilit: