On February 9, 2023, two significant earthquakes of magnitudes Mw 7.8 and Mw 7.7 struck southern Turkey, devastating the Kahramanmaraş region, causing nearly 50,000 fatalities, and displacing millions of people (Zhang et al., 2023). The first earthquake ruptured the Nurdağı–Pazarcık Fault, a subsidiary of the East Anatolian Fault Zone (EAFZ), while the second occurred about 90 km away on the Çardak–Sürgü Fault. This event later became known as the Kahramanmaraş earthquake doublet. With a combined magnitude of Mw 8, it represents the largest recorded continental strike-slip earthquake sequence.
The main strand of the EAFZ is a well-known seismically active zone, with many seismic records and studies conducted on the area. In contrast, the earthquake on the Çardak–Sürgü Fault was unexpected, as it had remained quiet for several centuries. This has left the Çardak–Sürgü Fault relatively understudied compared to the main strand of the EAFZ, particularly in terms of its kinematics and seismic risk.
In this study, we estimate the slip deficit along the main strand of the East Anatolian Fault Zone (EAFZ) and the Çardak–Sürgü Fault prior to the Kahramanmaraş earthquake doublet, using elastic strike-slip and dip-slip models combined with data assimilation. We find that the slip-deficit rates for the EAFZ are estimated at 2–8 mm/yr in the Anatolia-fixed frame and 0–11 mm/yr in the Arabia-fixed frame. For the Çardak–Sürgü Fault, slip-deficit rates are 1–4 mm/yr in the Anatolia-fixed frame and 2–11 mm/yr in the Arabia-fixed frame.
Dip-slip rates show wider variability. Along the EAFZ, they range from 0 to –12 mm/yr in the Anatolia-fixed frame and from –12 to 2 mm/yr in the Arabia-fixed frame. For the Çardak–Sürgü Fault, dip-slip rates range from –4 to 5.5 mm/yr in the Anatolia-fixed frame and from –3 to 10 mm/yr in the Arabia-fixed frame.
To further evaluate our estimates, we compared the interseismic slip deficit with coseismic slip distributions from the 2023 Kahramanmaraş earthquakes. The results indicate that the estimated slip deficit is broadly consistent with the observed coseismic slip for the Amanos and Çardak segments, but underestimates slip for the Pazarcık, Erkenek, and Göksun segments. A key limitation of the approach is that the locking depth remains poorly constrained in the joint inversion, particularly when using GNSS velocities in the Arabia-fixed frame rather than the Anatolia-fixed frame. In addition, the choice of kinematic representation for the joint fault mechanism exerts a strong influence on the results, leading to notable differences in the estimated slip deficit.