The Queen Maud Land (QML) region of Antarctica represents an anomaly in the context of global ice sheet dynamics. QML has exhibited significant mass gains in recent years in contrast to most other polar regions. Understanding the causes and consequences of this behaviour is the core objective of the EXQALIBR project (EXceptional Queen mAud Land Ice mass Balance incRease – causes and effects), which combines satellite observations, ice sheet modelling, and field campaigns at the Finnish Aboa research station. Within this project, space gravimetry from the GRACE and GRACE-FO missions provides a direct measure of total mass changes, complementing altimetric and input/output estimates derived in parallel workstreams of the project. However, the limited resolution of GRACE(-FO) prevents the estimation of local mass balance signals at the scale of individual drainage basins or precipitation events. This study presents an approach to enhance the spatial resolution of GRACE-derived gravity fields down to a few kilometers using Geographically Weighted Regression (GWR). This locally adaptive regression technique accounts for the spatial non-stationarity of the gravity signal by fitting spatially varying coefficients, allowing the integration of high-resolution ancillary data, such as ice elevation products, firn compaction models, and climate model outputs. Applied over the QML region, this approach yields high-resolution mass change maps that better resolve the spatial patterns of recent ice mass gain. These results are compared against other estimates within EXQALIBR to better constrain the mass balance components driving this Antarctic QML signal