Detection Of Sea Ice Leads In The Beaufort Sea Using Remote Sensing Imagery And Estimation Of Energy Budget Over Leads Surface

Sea ice is a basic element of Arctic climate,while sea ice leads are linear structures of ocean surface within pack ice that is exposed to the atmosphere during an opening event caused by various forces,such as wind and water stresses.However,in remote sensing of sea ice and leads,due to the limitation of image resolution and the presence of mixed pixels,thin ice in pack ice is also regarded as a certain stage of lead development,i.e.,refrozen leads.Sea ice leads break the insulation of sea ice between the Arctic Ocean and the atmosphere and thus,become the prime window for heat and vapor exchange in the frozen ice.As sea ice in the Arctic Ocean has experienced significant declines in both extent and thickness,a positive trend in the drift speed of sea ice had been observed in several studies.Earlier opening and retreat dates in the Beaufort Sea related to increased easterly wind,driven by enhanced Beaufort High(BH)were reported.Based on their results,along with the loss of sea ice,especially for multi-year ice,and the inter-annual change in atmosphere circulation pattern,variations in seasonal cycle/pattern and inter-annual trend of sea ice leads distribution can be expected.With the acceleration of sea ice areal reduction and thinning in the Arctic,it is of great importance to study the spatial-temporal distribution of ice leads and its impact on surface energy budget,and sequentially basal and literal melt of sea ice through local thermal process.The thesis is designated to contribute to the subject in three aspects:a.New algorithm for detection of sea ice leads from satellite thermal images.This study presents a modified algorithm using multiple thresholds and daily scores to retrieve spring leads in the Beaufort Sea using temperature anomaly obtained from Terra MODIS thermal images.Both the magnitude and persistence of temperature anomaly are taken into account through the use of multiple thresholds and accumulated score.To mitigate the influence of rapid ice motion,only MODIS thermal images from successive Terra orbits within a day defined by local the time at 135°W were used to generate a daily lead map for the month of April,from 2001 to 2020.The development of April leads in the Beaufort Sea were identified in 2019.Lead area increased steadily by 1 794 km² per day on average,except for the cloudy days at the end of the month.Reanalysis data from ERA-interim shows that the air temperature in the study area experienced a multistep increase by about 0.65k/day since 9 April.The easterly wind was strong and consistent in the middle of the month,driving a fast westward ice motion.As a result,the maximum width of sea ice leads in the Beaufort Sea nearly doubled towards the end of the month.The monthly mean lead area and the first 10-day average for April were calculated using images from 2001 to 2020.Corresponding to the warming shown by air temperature,a rising trend in lead area was identified in the April mean series,estimated at 2 612±1 245 km² per year.The change rate for early April alone is higher compared to the monthly average and was estimated at 3 383±1 628 km² per year.For the two groups of selected days with cloud cover of0.3 and 0.5,the rates of increase were about 26.4?29.2 per cents lower than that from the 10-days mean in early April.We found most of the positive anomalies in the detected lead area and orientation can be explained by enhanced BH,persistent easterly wind,and high drift speed of sea ice.Detected lead area also presents a close correlation with the 11-day average of ice drift speed prior to lead detection.Our results also show the influence of regional pressure gradient and easterly wind on the lead area is limited to the synoptic scale and recedes within a week.Compared to existing lead datasets,our method was able to capture more of the narrow leads and detected larger lead area than Willmes and Heinemann(2015a)and Hoffman et al.(2019).For the overlapping period,our result gave the slowest increasing trend in lead area on the base of multi-year mean,rising about 3.3%per year for April mean and 4.5%for early April,similar to the W&H data,but much lower than the Hoffman dataset.b.Estimation of turbulent heat flux over leads using satellite thermal imagesAlthough the same local temperature anomaly and threshold method were applied,leads retrieved at MODIS and Landsat-8 TIRS resolution scales present very different geometry and distribution.Within the studied area,total length of leads is 10150.3 km from TIRS,including 8502.2 km(83.76%)from small leads with width less than 1 km.This is in contrast to the total length of 2746.4 km from MODIS,where the narrow leads(1km wide)only account for 1050.0 km(38.23%).Total length of leads is underestimated by 72.9%in MODIS data.For the area of leads,small leads(width?1km)account for 34.54%of the total lead area from TIRS,while only take up 13.00%of the total lead area in MODIS.Although lead widths follow the power low distribution at both scales,the fitted exponents vary from 2.241 to 2.346.When bulk aerodynamic formulae were applied with the reanalysis dataset,heat flux estimated using TIRS data is 8.40×10¹¹ W,56.70%larger than that from MODIS data(5.36×10¹¹ W).Small leads account for 2.16×10¹¹ W(25.75%)of total heat flux over all leads in TIRS data,almost seven times the heat flux from the narrow lead category in MODIS(3.10×10¹⁰W,5.79%).Turbulent heat flux over leads estimated from TIRS data by Andreas and Cash(1999)model is 1.11×10¹² W,32.34%higher than that from bulk formulae(8.40×10¹¹W).In both cases,small leads account for about a quarter of total heat flux in both models,due to its large area,though the heat flux estimated using fetch-limited model is 41.39%larger.More contribution from small leads can be expected at larger temperature difference and stronger wind conditions.c.Influence of lead distribution on regional surface radiation budgetThe temporal-spatial distribution of surface albedo,cloud cover,surface net radiation and cloud radiative force of the Arctic Ocean is demonstrated in the thesis based on albedo and cloud fraction from the CLARA-A2 dataset and surface radiation estimation from the ERA-5 reanalysis dataset.Contribution from sea ice concentration and lead to the regional surface albedo in the Beaufort Sea is discussed.The surface albedo of the April leads derived from Terra/MODIS is fitted with a mean of 0.33,compared to 0.73 for the surrounding thick ice.Fitted net shortwave radiation for lead surface is 2.4 times of that from thicker ice,and about twice of longwave radiation for lead surface compared to ice surface.The net radiation of April lead surface in the Beaufort is about neutral,while for the thick ice surface,a mean of 8.5W/m² radiation is transferred from ice to the atmosphere.In-situ measurement of surface parameters is still needed for accurate estimation of energy budget over lead surfaces and validation of radiation and turbulent heat flux estimation from satellite remote sensing.