Wildfire management and response requires frequent and accurate burnedarea mapping. How to map daily burned areas with satisfactory accuracyremains challenging due to missed detections caused by accumulating activefire points as well as the low temporal resolution of sensors onboard satelliteslike Sentinel-2/Landsat-8/9 and monthly burned area product generatedfrom the Visible Infrared Imaging Radiometer Suite (VIIRS) data. ConvNetbasedand Transformer-based deep-learning models are widely applied tomid-spatial-resolution satellite images. But these models perform poorly onlow-spatial-resolution images. Also, cloud interference is one major issuewhen continuously monitoring the burned area. To improve detection accuracyand reduce cloud inference by combining temporal and spatial information,we propose an autoregressive spatial-temporal model AR-SwinUNETRto segment daily burned areas from VIIRS time-series. AR-SwinUNETRprocesses the image time-series as a 3D tensor but considers the temporalconnections between images in the time-series by applying an autoregressivemask in Swin-Transformer Block. The model is trained with 2017-2020wildfire events in the US and validated on 2021 US wildfire events. The quantitativeresults indicate AR-SwinUNETR can achieve a higher F1-Score thanbaseline deep learning models. The quantitative results of testset which consistsof eight 2023 long-duration wildfires in Canada show a better F1 Score(0.757) and IoU Score (0.607) than baseline accumulated VIIRS Active FireHotspots (0.715) and IoU Score (0.557) compared with labels generated fromSentinel-2 images. In conclusion, the proposed AR-SwinUNETR with VIIRSimage time-series can efficiently detect daily burned area providing better accuracy than direct burned area mapping with VIIRS active fire hotspots.Also, burned area mapping using VIIRS time-series and AR-SwinUNETRkeeps a high temporal resolution (daily) compared to other burned area mappingproducts. The qualitative results also show improvements in detectingburned areas with cloudy images.