The study was conducted in three districts of Tanzania, namely Lushoto, Pangani, and Bagamoyo (the latter two are coastal districts) ( Figure 1 ). Lushoto and Pangani districts are both located in the Tanga Region, while Bagamoyo District is found in the Pwani Region.

Geospatially, Bagamoyo District is located at 6˚26′31″ South and 38˚54′ 15″ East. The district has a humid tropical climate with seasonal average temperatures ranging from 13˚C - 30˚C. The annual rainfall in the district ranges between 800 and 1200 mm per annum, with the coastal strip receiving relatively more precipitation than the up-country. The main socio-economic activities conducted in the district include farming, livestock, tourism, beekeeping, fisheries, and petty businesses.

Kidomole village is amongst the four villages of Fukayosi ward in Bagamoyo district. The village has decreasing water catchments, including rivers and natural dams, which support small-scale farming, domestic uses, and livestock. The villagers are mainly engaged in small-scale agriculture practices, livestock keeping, petty business, riverine fisheries, and boda-boda business.

Lushoto District is situated in the northern part of Tanga Region within 4˚25' - 4˚55' latitude south of the Equator and 30˚10' - 38˚35' longitude east of Greenwich ( CAN Tanzania, 2021 ). A significant part of Lushoto District is within the Western Usambara Mountains, which lie between 300 and 2,100 meters above sea level. The main socio-economic activities conducted in the district include farming, livestock, tourism, beekeeping, and petty businesses. Mwangoi village is one of the two villages found in Mwangoi ward, Lushoto District. Climate change-related hazards and risk levels, particularly those related to droughts, are high in the village, severely impacting the community overall.

Pangani District a region lies between 5˚15' to 6˚ South of the equator and 38˚35' to 39˚ East of the Greenwich Meridian. The district has a humid tropical climate with average temperatures ranging from 24˚C to 33˚C. May to July is the coolest season, while December to February is the hottest. The district receives an average rainfall between 600 mm and 1200 mm per year, with more rain in the interior areas ( CAN Tanzania, 2021 ). Pangani District practices both commercial and subsistence farming, fisheries, and forestry as the main sources of income and livehood activities. Hunting, livestock keeping, trade and commerce, mining, and quarrying are some of the additional socio-economic activities in the district.

Ushongo village is located along the coastline of Pangani district, where communities are mainly engaged in various livelihoods, including fisheries, animal husbandry, petty business, and the farming of staple crops such as maize and beans, as well as commercial crops, including cashew nuts and coconuts. The village faces severe impacts from sea level rise, which has led to saltwater intrusion into freshwater aquifers, coastal erosion, and inundation of farm fields. The impacts of climate change on groundwater resources in coastal areas were reported in IPCC (2022) , with salinity being cited as one of the most notable impacts in coastal areas.

Three different composite indices were used to assess community vulnerability to climate change impacts in the three villages. The indices used are the Livelihood Vulnerability Index (LVI), the Livelihood Vulnerability Index-Intergovernmental Panel on Climate Change (LVI-IPCC) ( Small-Lorenz et al., 2013 ), and the Livelihood Effect Index (LEI). The LEI is constructed from the Sustainable Livelihood Framework to understand community vulnerability with respect to their livelihood strategies. Madhuri et al. (2015) used the LVI to assess the impact of floods depending on differences in livelihood in Bhagalpur district, India. The index was adapted to Tanzania to identify vulnerability and context-specific resilience measures.

The study applied a balance-weighted approach to calculate the Livelihood Vulnerability Index (LVI). Such computation takes into consideration external and internal factors relating to community vulnerability to climate change ( Young et al., 2011 ; Shah et al., 2013 ). Thus, essential parameters in LVI calculation include the Socio-Demographic Profile (SDP), Livelihood Strategies (LS), Finance (Fi), Knowledge and Skills (KS), Social Networks (SN), Health (H), Food (F), Water (W), and Natural Vulnerability and Climate Variability (NVCV). Due to the difference in measurement scales in each subcomponent, it was imperative to standardize each subcomponent using Equation (1) ( IPCC, 2022 ).

${\text{Index}}_{{\text{S}}_{\text{v}}}=\frac{{\text{S}}_{\text{v}}-{\text{S}}_{\min }}{{\text{S}}_{\max }-{\text{S}}_{\min }}$ (1)

where Sv is the observed initial value of the subcomponent for village v, and S_min and S_max are the minimum and maximum values, respectively.

After each subcomponent had been standardized, values were averaged using Equation (2) ( IPCC, 2022 ).

${\text{M}}_{\text{s}}=\frac{{{\displaystyle \sum }}_{\text{i}=1}^{\text{n}}{\text{Index}}_{\text{Svi}}}{\text{n}}$ (2)

LVI calculation for each village using inputs from Equation (2) on each of the major components was averaged using Equation (3) ( IPCC, 2022 ).

${\text{LVI}}_{\text{v}}=\frac{{{\displaystyle \sum }}_{\text{i}=1}^{\text{n}}{\text{W}}_{\text{Mi}}{\text{M}}_{\text{vi}}}{{{\displaystyle \sum }}_{\text{i}=1}^{\text{n}}{\text{W}}_{\text{Mi}}}$ (3)

where LVI_v is a livelihood vulnerability index for the village v, obtained by averaging the number of significant components. The weights of each major element, WMi, were obtained by considering. the number of subcomponents in each principal component so that the proportional contribution of each major element to the overall LVI is considered. The LVI is scaled from 0 (least vulnerable) to 1 (most vulnerable).

Exposure, adaptive capacity, and sensitivity are the three factors contributing to vulnerability to climate change. Calculating the Livelihood Vulnerability Index-IPCC (LVI-IPCC) deviates from LVI when the major components are combined rather than being merged in the early steps. The vulnerability contributing factor, CF, is calculated from Equation (4) ( IPCC, 2022 ).

$\text{CF}=\frac{{{\displaystyle \sum }}_{\text{i}=1}^{\text{n}}{\text{W}}_{\text{Mi}}{\text{M}}_{\text{vi}}}{{{\displaystyle \sum }}_{\text{i}=1}^{\text{n}}{\text{W}}_{\text{Mi}}}$ (4)

The value of adaptive capacity is calculated from the inverse of the subcomponents making up this factor. This is because the higher the adaptive capacity, the lower the vulnerability and vice versa. Once the three contributing factors were calculated, the LVI-IPCC was calculated from Equation (5) ( IPCC, 2022 ).

$\text{LVI}-{\text{IPCC}}_{\text{v}}=\left({\text{e}}_{\text{v}}-{\text{a}}_{\text{v}}\right){\text{S}}_{\text{v}}$ (5)

The Calculating Livelihood Effect Index (LEI) is derived from the livelihood capitals of the sustainable livelihood framework (SLF), which are natural, physical, human, social, and financial capitals. LEI is a household composite index of vulnerability. To calculate the LEI, the LVI values of major components were used to calculate the score values (C_v) for each livelihood capital (L_i) by combining, where n is the number of subcomponents forming the livelihood capital. LEI is then calculated as a weighted average of all livelihood capitals using Equation (6) ( IPCC, 2022 ).

${\text{C}}_{\text{v}}=\frac{{{\displaystyle \sum }}_{\text{i}=1}^{\text{n}}{\text{L}}_{\text{i}}}{\text{n}}$ (6)

$LEI=\frac{{{\displaystyle \sum }}_{i=1}^n{W}_i{C}_{vi}}{\sum W_i}$ (7)

Key informant interviews, focus group discussions, and structured household surveys were the main methods used to collect input data for the three indices. The participatory geographical information system was used for hazard and resource mapping, depicting the most vulnerable areas and their related hazards. The QGIS software was used for spatial representation of the climate hazards and vulnerabilities and visual interpretation of geospatial data. Climate change risk identification and prioritization were done by identifying climate change hazards and their associated risks. Climate change risks were later ranked based on existing knowledge, using stakeholders’ workshops and participatory risk mapping exercises ( Maharjan et al., 2017 ). The risk vulnerability was attained through ranking of risk level and severity (5: very high to 1: minimal) and frequency of occurrence (5: likely to occur annually and 1: likely to occur once per decade).

The climate risks for Mwangoi village are presented in Table 1 and the distribution of resources in Figure 2 and Figure 3 . Crop and livestock diseases and pests are the most serious climate-related risk events in the village, affecting crop and livestock productivity, thus affecting household incomes as well. An increased infestation of tomato bacteria, causing serious wilting, has been reported to seriously affect the income accruing from tomato sales. This usually happens when there is more than normal rainfall in the area. Hypo-calcium in cattle has been a serious issue as well, lowering livestock productivity. Moreover, land subsidence and water table rise have been recurring hazards due to frequent flooding events. Water shortage is another climate-related hazard in the Mwangoiarea. Water scarcity is more likely to increase household vulnerability to waterborne diseases and water-related conflicts during the dry season. As water is sourced mainly by women and children, the task reduces time that would have been used to attend to other economic activities and attend schools for children, especially girls. The task of sourcing water becomes even more stressful during the dry season.

The climate risks for Ushongo village are presented in Table 2 . Paddy farming is significantly affected by frequent dry spells due to limited water and/or sea level rise. Paddy is a hydrophytic plant; therefore, a limited water supply affects crop growth and yield. However, the hazard risk is at a medium level, scoring 3.7. Water quality deterioration due to sea level rise resulting from saltwater intrusion into freshwater aquifers, including existing hand-dug wells, is severe in the coastal village of Ushongo in Pangani district. The same applies to farm fields and thus makes the soil infertile for farming. Saltwater intrusion results in water scarcity for domestic use and animal watering and therefore poses health threats to this community and the neighboring one.

Floods occasionally inundate homes and damage road infrastructure. This hazard, however, is not as severe as sea level rise and storm surges, which have led to household migration due to sea inundation from severe coastal erosion ( Figure 4 ). Drought, despite being a spatially distributed hazard, has little impact on Ushongo village or Pangani district due to the nature of the area’s livelihood strategies. Furthermore, sea level rise and storm surges have had a greater impact on houses near the beach than on fishing activities. As a result, the village is shrinking, with the greatest push coming from the east, on the coast of the Indian Ocean.

On the western part, there is a fixed boundary, bordering the Mwera sisal estate. The risk of shrinking and ultimately disappearing is imminent and likely. Some fishing grounds have been pushed farther by sporadic sea level rise. This has heightened the difficulty of getting the once easily appropriable fish catch. Some families have been pushed inland by seawater, with one family already forced to vacate and abandon the house due to irreparable damage. Figure 5 shows the distribution of resources at Ushongo village in Pangani district.

Moreover, some households are engaged in petty trade, a secondary livelihood activity based on the income from fisheries practices. Therefore, any impact on fisheries undertakings due to climate change and variability is likely to sporadically affect petty businesses in the village. The impact of climate change on groundwater resources in coastal areas was reported in IPCC (2022) , and specifically, the increase in salinity was cited as one of its most notable impacts. Therefore, the situation in Ushongo reiterates what has already been reported.

The climate risks for Mwangoi village are presented in Table 3 . Kidomole village is the least vulnerable of all three villages. Drought, with the likelihood of occurring once in 5 years, affects crop yield and income from the sale of crops. Water shortages and floods are the most serious climate risks in Kidomole village. Prolonged dry spells lead to low crop yields, affecting the incomes of smallholder farmers in the village. Droughts also led to the death of cattle due to limited water and grazing areas. Floods damage homes, properties, and agricultural land, displacing some families.

No serious incidents of livestock diseases and pests were reported. Generally, prolonged dry spells are the most troublesome climate risks ( Figure 6 ), followed by heavy rainfall that causes sporadic floods. Low crop yield and loss of income are the most serious risks due to drought in the village. The incidences of livestock diseases and pests are very limited, and thus the risk they pose is insignificant (1.7). However, the damage to houses due to floods is a climate risk that needs serious attention. If left unattended, it can lead to precarious damage in the future. The distribution of resources at Kidomole village in Bagamoyo district is presented in Figure 7 .

The findings from Table 4 indicate LVI results of nine primary components in Mwangoi, Kidomole, and Ushongo villages. The overall LVI in the study villages ranges from low in Ushongo to moderate in Kidomole and Mwangoi. Moreover, some of the major components have varying vulnerability indices. For example, social network components have the highest vulnerability in all three villages (0.68 in Kidomole, 0.69 in Ushongo, and 0.73 in Mwangoi). This indicates that in these villages, weak social ties may increase vulnerability to climate change risks, shocks, and stresses.

The results also reveal that few individuals have either assisted or borrowed money from their neighbours, and a substantial percentage (more than 70% in each village) demonstrate a lack of affiliation to any financial organisation, access, or networking. This increases vulnerability to climate change impacts as reported by Thomas et al. (2019) . Lack of affiliation with any borrowing or lending organization, such as Village Community Banks (VICOBA), affects individual and community adaptive capacity, thus decreasing their ability to cope with climate change shocks.

Another significant component with moderate to highest vulnerability in all villages is access to food, with 0.603 in Kidomole, 0.523 in Mwangoi, and 0.480 in Ushongo. High vulnerability in Kidomole may be due to inadequate crop diversification, as the crop diversification index (CDI) is 0.5, which is higher than in Ushongo (CDI = 0.3) and Mwangoi (CDI = 0.4). The CDI indicates that the variety of crops cultivated in Kidomole is less compared to Ushongo and Mwangoi. Furthermore, the findings suggest that in Mwangoi and Kidomole villages the most important crops are maize and beans, which are rain-fed and therefore highly susceptible to the impacts of climate change.

Moreover, horticultural crops such as tomato, cabbage, and spinach are dominant under irrigation in the study villages. In Mwangoi village, the respondents revealed that they own the land under cultivation and obtain water from rivers in both dry and wet seasons. Similarly, in Kidomole village, maize and cassava are dominant crops in the rainy season. However, per household, crop diversity is moderate, with other crops such as banana, rice, and groundnuts equally cultivated in Kidomole. Conceivably, crop diversification improves capacity, reducing vulnerability ( Kihila, 2018 ). Climate problems affecting one crop might not affect another crop equally; therefore, planting two diverse crop varieties may result in better productivity than a single crop or two crops with similar physiological needs.

The Livelihood Effect Index in Figure 9 shows the effects of climate change for each type of capital or resource at the household level in the case study villages. The Livelihood Effect Index considered five capitals, calculating an index for each resource in each village. The results from this index (LEI) suggest a low livelihood vulnerability to climate change, with overall values of 0.410 for Mwangoi village, 0.414 for Kidomole, and 0.375 for Ushongo (0 indicating low vulnerability and 1 indicating high vulnerability). However, some effect dimensions have moderate vulnerability across the villages, but when the dimensions are aggregated, the overall vulnerability decreases (i.e., low). For example, the health dimension under human capital has moderate vulnerability in both Kidomole and Ushongo villages, and Mwangoi has moderate vulnerability in the socio-demographic dimension. Results from the LEI are slightly higher than results from the LVI for Ushongo village, but other village values remain similar.