Ionospheric effects represent a significant source of error in Interferometric Synthetic Aperture Radar (InSAR) measurements, especially for long-wavelength sensors such as ALOS-2 PALSAR-2. This study investigates the accuracy of estimating ionospheric phase using the Split Spectrum Method (SSM) in a time series of L-band SAR acquisitions spanning 2020–2021 in an adverse environment that limits the performance of SSM. The estimated ionospheric phase is evaluated by calculating the phase closure of differential ionospheric phase triplets and averaging them in space for each date. Additionally, estimates are compared with Vertical Total Electron Content (VTEC) maps from two external sources: the International GNSS Service (IGS) Global Ionosphere Maps (GIMs) and the Madrigal upper atmospheric science database.

Results show high phase closure values for dates that, due to low coherence, do not form many interferometric pairs. A strong correlation (r=0.82) is observed between InSAR-derived and Madrigal-based ionospheric phase, despite data gaps in the latter. In contrast, low correlation is found with the IGS VTEC maps, attributed to their coarse spatial resolution and high modeled ionospheric shell height. The linear correlation between SSM and Madrigal data demonstrates good agreement, highlighting the potential of integrating external VTEC maps as complementary information, particularly in cases of low SAR coherence or limited acquisitions where the SSM becomes less effective.