Validation of snow extent time series derived from AVHRR GACdata at Himalaya-Hindukush

Abstract. Long-term monitoring of snow cover is crucial for climate and hydrology studies. To meet the increasing demand for a long-term, consistent snow product, an exceptional snow cover climatology was generated dating back to the 1980s using AVHRR GAC data. However, the retrieval of snow extent is not straightforward due to artifacts introduced during data processing, which are partly caused by the coarse spatial resolution of AVHRR GAC data, but also heterogeneous land cover/topography. Therefore, the accuracy and consistency of this long-term AVHRR GAC snow cover climatology needs to be carefully evaluated prior to its application. Here, we extensively validate the AVHRR GAC snow cover extent dataset for the Hindu Kush Himalaya (HKH) region. The mountainous HKH region is of high importance for climate change, impact and adaptation studies. Additionally, the influences of snow depth, land cover type, elevation, slope, aspect, and topographical variability, as well as the sensor-to-sensor consistency have been explored using a snow dataset based on long-term in situ stations and high-resolution Landsat TM data. Moreover, the performance of the AVHRR GAC snow cover dataset was also compared to that of MODIS (MOD10A1 V006). Our analysis shows an overall accuracy of 94 % in comparison with in situ station data. Using a ±3 days temporal filter caused a slight decrease in accuracy (from 94 to 92 %), which is still comparable to MOD10A1 V006 (93.6 %). Validation against Landsat5 TM data over region of P140-R40/41 indicated overall RMSEs of about 13 % and 16 % and overall Biases of about −1 % and −2 % for the AVHRR GAC raw and gap-filled snow datasets, respectively. It can be concluded that the here validated AVHRR GAC snow cover climatology is a highly valuable and powerful dataset to assess environmental changes in the HKH due to its good quality, unique temporal coverage (1982–2018), and inter-sensor/satellite consistency.