How to adjust Dark Sky API response data for daylight savings (GMT/BST)

Question

I'm very new to python and having trouble adjusting the results of an api request to handle UK daylight savings (Greenwich Mean Time / British Summer Time). When reading the Dark Sky documents they state :

The timezone is only used for determining the time of the request; the response will always be relative to the local time zone

I have built the below code to return historic weather information based upon hourly min/max/avg temperatures for each day for specific weather stations. I've been able to turn the returned UNIX time into a time stamp, but I need a way to deal with the data when the clocks change.

This is my whole code, if anyone can offer any advice I would be very grateful

import requests
import json
import pytemperature
import pandas as pd
pd.options.mode.chained_assignment = None  # default='warn'

from pandas.io.json import json_normalize
import datetime

#Import CSV with station data
data_csv = pd.read_csv("Degree Days Stations.csv") 

#Basic Information for the API request
URL = "https://api.darksky.net/forecast/"
AUTH = "REDACTED/"
EXCLUDES = "?exclude=currently,daily,minutely,alerts,flags"

#Use ZIP function to loop through the station_df dataframe
for STATION, NAME, CODE, LAT, LON in zip(data_csv['Station'], data_csv['Name'], data_csv['Code'], data_csv['Lat'], data_csv['Lon']):

    #Result is based upon the time zone when running the script
    date1 = '2019-10-26T00:00:00'
    date2 = '2019-10-29T00:00:00'

    start = datetime.datetime.strptime(date1, '%Y-%m-%dT%H:%M:%S')
    end = datetime.datetime.strptime(date2, '%Y-%m-%dT%H:%M:%S')
    step = datetime.timedelta(days=1)
    while start <= end:
        #Compile the Daily Values
        DATE =  (datetime.date.strftime(start.date(), "%Y-%m-%dT%H:%M:%S"))

        #build the api url request
        response = requests.get(URL+AUTH+str(LAT)+","+str(LON)+","+str(DATE)+EXCLUDES+"?units=si")
        json_data = response.json()

        #Flatten the data 
        json_df = json_normalize(json_data['hourly'],record_path=['data'],sep="_")
        #Extract to new df
        output_df = json_df[['time','temperature']]

        #insert station name to dataframe for my debugging
        output_df.insert(0, 'Name', NAME)


        #Convert UNIX date to datetime
        output_df['time'] =  pd.to_datetime(output_df['time'], unit = 's')
        ##################
        # Deal with timezone of output_df['time'] here 
        ##################


        #Convert temperatures from oF to oC
        output_df['temperature'] = pytemperature.f2c(output_df['temperature'])
        #Find the MIN/MAX/AVG from the hourly data for this day
        MIN = output_df['temperature'].min()
        MAX = output_df['temperature'].max()
        AVG = output_df['temperature'].mean()

        #Build the POST query
        knackURL = "https://api.knack.com/v1/objects/object_34/records"
        payload = '{ \
                   "field_647": "' + STATION + '", \
                   "field_649": "' + NAME + '", \
                   "field_650": "' + str(CODE) + '", \
                   "field_651": "' + str(DATE) + '", \
                   "field_652": "' + str(MIN) + '", \
                   "field_653": "' + str(MAX) + '", \
                   "field_655": "' + str(AVG) + '" \
                  }'

        knackHEADERS = {
               'X-Knack-Application-Id': "REDCATED",
               'X-Knack-REST-API-Key': "REDACTED",
               'Content-Type': "application/json"
               }
        #response = requests.request("POST", knackURL, data=payload, headers=knackHEADERS)
        start += step

My results for October 27th (BST) and 28th (GMT) are shown below and are relevant to the current timezone (GMT). How can I ensure that that I get the same size of dataset starting from 00:00:00 ?

I've looked at arrow and pytz but can't seem to get it to work in the context of the dataframe. I've been working on the assumption that I need to test and deal withe the data when converting it from Unix... but I just can't get it right. Even trying to pass the data through arrow.get eg

 d1 = arrow.get(output_df['time'])
 print(d1)

Leaves me with the error : "Can't parse single argument type of '{}'".format(type(arg)) TypeError: Can't parse single argument type of ''

So I'm guessing that it doesn't want to work as part of a dataframe ? Thank you in advance