mysql - Optimal Way to Setup Table Structure -

i designing new mysql backend system , faced issue , unsure of optimal way setup table. let's have table called cities. each city has relation table weather contains weather information every month. simple let's structure of cities table is:

--------------- | id |  name  | --------------- |  1 | city1  | |  2 | city2  | |  3 | city2  | --------------- 

for weather, have approximately 50 fields. each month of year have high temp, low temp, avg temp, avg precipitation. having table 50 columns seems inefficient , seems difficult add new fields. other option can consider have table following setup:

---------------------------------------------------- | id | cityid |    type   | value  | unit |  month | ---------------------------------------------------- |  1 |    1   | high_temp |   50   |  f   |    1   | |  2 |    1   | low_temp  |   35   |  f   |    1   | |  3 |    1   | avg_temp  |   45   |  f   |    1   | |  4 |    1   | avg_prec  |   10   |  in  |    1   | |  5 |    2   | high_temp |   60   |  f   |    1   | |  6 |    2   | low_temp  |   50   |  f   |    1   | |  7 |    2   | avg_temp  |   55   |  f   |    1   | |  8 |    2   | avg_prec  |   2    |  in  |    1   | ---------------------------------------------------- 

however, have around 1000 cities in database , extending tens of thousands. having table around 5 million rows seems excessive , worry query time.

what optimal way set up, one-to-one relationship or one-to-many relationship? also, more efficient creating complex queries in try search cities based on multiple weather characteristics?

currently have shown not normalized structure. imagine questions like:

what cities high temp greater 30 , low temp less 15 in january?

for city 1 when max temp , min temp exceeded?

query complicated ones. , questions involving several types of temp give headache in future.

the optimal way move temp values columns:

id | cityid |  hightemp | lowtemp | avgtemp | avgprec | unit |  month | 

now abovemensioned query simple select statement appropriate clause. need ask similar questions , think how whould answer in concrete schema.

and believe, 1 millions of rows nothing database engine appropriate indexing.