The review of the physical environment described by the authors in this section led to suggestions of further research which can reduce remaining uncertainties; these recommendations are summarized below while recommendations that are particularly important for improving Arctic NEBA are listed separately. 

1. Seasonal and interannual fluctuations. The prevailing environmental characteristics such as extreme cold temperatures, variable solar irradiation and ice coverage affect the marine ecosystem and its populations. Seasonal productivity ties in with the degree of solar irradiation and ice coverage and directly affects species population recruitment and migration patterns.
   1. A better understanding of ice, and the corresponding variance of faunal distributions and life-history patterns on a spatial (regionally and throughout the pan-artic region) and temporal (annual and interannual) scale would be especially useful to protecting areas of densely populated VECs. 
2. Interface habitats. The physical structure of the Arctic provides additional interface as well as open water environments that are occupied by VECs.
   1. Interface environments that are stable (e.g. shorelines, sediment/water interface, edges of multi-year ice, convergence of water masses near subsurface topographic features) have generally been mapped throughout most of the Arctic. Where they have not been mapped they need to be. Seasonal use patterns by VECs need to be summarized by these interface ECs (see Section 2). 
   2. Interface environments that are variable (annual ice edges including polynyas and under-ice brine channels, fresh water releases and their transport, barrier island beaches/shorelines should be characterized by indicating most likely locations in the Arctic for these less permanent features. Understanding of patterns of use by VECs will be strengthened under work proposed in Section 2. 
3. Alterations in chemical processes. We need to better understand how cold temperatures and extreme weather interacts with broken and more permanent ice to affect the chemical processes associated with the fate of oil compounds and natural biodegradation rates (Sections 3 and 5). 

The recommendations presented below indicate where increased knowledge of physical environment would result in reducing existing uncertainties in NEBA assessments. No prioritization has been made to the list; for some of the recommendations, surrogate data may be already available. 

1. Arctic ecosystem consequence analyses need to include impact assessments of open pelagic waters as well as shorelines, river discharges, sediment/water interfaces, ice/water interfaces, convergence zones and surface layers near the air/water interface. These ECs may receive concentrated oil exposure depending on the type of OSR actions employed, and should receive further study.
   1. The location of primary ECs used by VECs could be added to regional sensitivity maps already available. 
   2. The locations where less permanent ECs can be prevalent are also important to VEC distribution and need to be located generally so that site specific assessments of their presence can be added to environmental sensitivity maps for reference during response actions. The environmental features that control the locations of these less permanent features and means to determine their presence during a response action are an area of additional study.