Ais update

docs/_toc.yml

@@ -6,6 +6,7 @@ parts:
     chapters:
       - file: ais/ais_intro
       - file: ais/ais_trade
+      - file: ais/ais_hhi
       - file: ais/ais_emissions
       - file: ais/ais_fishing
   - caption: Tourism

docs/ais/ais_hhi.md

@@ -0,0 +1,13 @@
+# Market concentration of vessel operators
+Market concentration among vessel operators in Pacific Island countries can be analyzed using the Herfindahl-Hirschman Index (HHI):
+$$
+HHI =  \sum s_i^2 
+$$
+
+Where $s_i$ is the  market share of operator $i$. HHI value close to 0 indicates near perfect competition, while HHI value close to 1 indicates monopoly. 
+
+Here, the market share is defined by the share of port calls from an operator. We use the [port calls data](ais_trade.md) derived from AIS, and the operator information from ship register data. Note that this excludes all vessels not registered with the IMO which is ~12% of port calls from cargo, tanker, and passenger vessels. 
+
+## Yearly HHI Per Country per Vessel Category 
+
+% suggest 1 graph per vessel category (cargo, tanker, passenger)

docs/ais/ais_intro.md

@@ -25,6 +25,7 @@ This interactive map contains the following layers:
 ## Applications 
 
 ### [Port Arrivals and Trade Volume](ais_trade.md)
+### [Operator Market Concetration](ais_hhi.md)
 ### [Greenhouse Gas Emissions from Vessels](ais_emissions.md)
 ### [Fishing Intensity (Global Fishing Watch)](ais_fishing.md)
 

docs/ais/ais_trade.md

@@ -1,19 +1,23 @@
 # Port Arrivals and Trade Volume
 
-This page showcases results from our paper **Estimating Trade Volume in the Pacific Islands using Automatic Identification System (AIS)**, where we validate the use of AIS data to produce port-level statistics for the Pacific Islands. The paper is available [link]. 
+This page presents AIS-derived port-level statistics for the Pacific Islands. Motivated by the work of {cite:t}`Arslanalp2021`, our initial study focus on port arrivals and trade volume estimates for Samoa {cite:p}`WBAIS-Samoa`, and to the rest of the pacific countries {cite:p}`WBAIS-Pacific`. We find that our derived port calls data accurately capture international trade-related ships, and whilst cargo volume levels are off from official data, they can still capture variation when pooled together. Here, we present an updated version of these statistics, incorporating AIS data from as early as 2011 providing a longitudinal dataset for further study.
 
-This study covers Fiji, Kiribati, Marshall Islands, Micronesia, Nauru, Palau, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu, and Vanuatu. The primary data sources are AIS data and ship registry from the UN Global Platform (UNGP), and global port boundaries from the {cite:t}`Arslanalp2021` study. The period covered by this study is from January 2019 to April 2023.
+## Data and Methods
+The datasets used are hourly AIS data within the pacific region from 2011-2024 provided by S&P, ship register data from the UNGP, port boundaries from {cite:t}`Arslanalp2021`. The countries covered are Fiji, Kiribati, Marshall Islands, Micronesia, Nauru, Palau, Papua New Guinea, Samoa, Solomon Islands, Tonga, Tuvalu, and Vanuatu. We followed the movement aggregation method from the {cite:t}`ADB2023` to capture the port arrivals making use of their helper functions available in the `ais` python package.    
 
-We follow two existing methodologies on trade volume estimation ({cite:t}`Arslanalp2021`; {cite:t}`Jia2019`) to estimate the cargo carried by each vessel upon arrival and departure. These papers utilize dynamic information on ship movements, static characteristics of each ship, and reported draft (depth of submergence), to estimate the amount of goods offloaded or loaded at a certain port. We find that our derived port calls data accurately capture international trade-related ships, and whilst cargo volume levels are off from official data, they can still capture variation across ports and relative trends within each port.
+For trade volume estimation, we follow the methods from {cite:t}`Arslanalp2021` which is based on the volume displacement of the ship. To get the volume of the cargo, the vessel's displacement upon arrival and departure are estimated using the formula:
 
-## Map of Ports
+$$
+Disp_d = L \times W \times d \times \rho \times c_d 
+$$ (dispeq)
 
-The map below shows the location of each port and the buffer area (22 km) used to extract AIS data.
+Where $Disp_d$ is the vessel displacement given length $L$, width $W$, reported draft $d$, density of salt water $\rho$ and block coefficient at reported draught $c_d$. To get $c_d$, the block coeffient at maximum levels, called *design* block coefficient $c_D$ is required. Here, we introduce a derivation of $c_D$ from {eq}`dispeq` by using maximum draught $D$ and maximum displacement $Disp_D$ which are availabe in the ship register data: 
 
-<div id="content" style="max-width: 100%; position: relative; padding-bottom: 56.25%; height: 0; overflow: hidden;">
-  <iframe src="../interactive/ais/PacificIslandsMap.html" name="Pacific Islands Map" id="Pacific Islands Map" style="border: 0; position: absolute; top: 0; left: 0; width: 100%; height: 100%;" allowfullscreen="">
-  </iframe>
-</div>
+$$
+c_D = \frac{Disp_D}{L \times W \times D \times \rho } 
+$$ (cdesign)
+
+We get the difference between the departure and arrival displacements to get the estimate of volume of cargo loaded/unloaded. If the net displacement is positive, it is assumed that the cargo for exports, and for imports otherwise. 
 
 ## Port Arrivals 
 
@@ -23,11 +27,13 @@ The map below shows the location of each port and the buffer area (22 km) used t
 
 <div class="flourish-embed flourish-chart" data-src="visualisation/19837444?2274258"><script src="https://public.flourish.studio/resources/embed.js"></script><noscript><img src="https://public.flourish.studio/visualisation/19837444/thumbnail" width="100%" alt="chart visualization" /></noscript></div>
 
+
 ## Data Availability
 
 The output data from this analysis is publicly available through the [Development Data Catalog](https://datacatalog.worldbank.org/search/dataset/0062856/Pacific%20Observatory%20Datasets?version=6).
 
 ## References
 
 ```{bibliography}
+:filter: docname in docnames
 ```

docs/references.bib

@@ -327,3 +327,26 @@ @INPROCEEDINGS{Dirk2014
  number={},
  pages={1457-1460},
  doi={10.1109/IGARSS.2014.6946711}}
+
+@article{ADB2023,
+      author = {{Asian Development Bank}},
+      title = "Methodological Framework for Unlocking Maritime Insights Using
+               Automatic Identification System Data",
+      journal = "Key Indicators for Asia and the Pacific",
+      year = "2023",
+      publisher = "Asian Development Bank",
+      isbn = "978-92-9270-367-7",
+      doi = "10.22617/FLS230401-2"
+}
+
+@unpublished{WBAIS-Samoa,
+  author = {Chico, Cherryl and Chamorro, Andres},
+  title = {Samoa Port Statistics Using AIS},
+  note = {unpublished}
+}
+
+@unpublished{WBAIS-Pacific,
+  author = {Chico, Cherryl and Chamorro, Andres},
+  title = {Estimating Trade Volume in Pacific Islands using AIS},
+  note = {unpublished}
+}